India - National Water Management Project (Vol. 1 of 3) : Volume one

. Document of The World Bank FOR OFFICIAL USE ONLY Report No. 6468-IN STAFF APPRAISAL REPORT INDIA NATIONAL WATER MANAGEMENT PROJECT February 20, 1987 South Asia Projects Department Irrigation II Division This document has a restricted distribution and may be used bs recipients only in the performance of their official dutie%- It, enntPntc m, nnt ntfhrs ;C h. Ai.-In.-A ' UA'It W  CURRENCY EQUIVALENTS US$1.00 5 Rupees (Rs) 13.0 WEIGHTS AND MEASURES (METRIC SYSTEM) 1/ 1 meter (m) = 3.28 feet (ft) 1 kilometer (km) = 0.62 miles (mi) 1 hectare (ha) = 2.47 acres (ac) 1 million cubic meters (Mm3) = 804 acre-feet (ac-ft) 1 cubic foot per second (cfs or cusec) = 0.0283 cubic meters per second (m3/s) 1 kilogram (kg) a 2.2 pounds (lb) 1 metric ton (t) = 2,205 pounds (lb) 1 thousand million cubic feet (TMC) = 28.317 million cubic meters (Mm3) FISCAL YEAR 0OI, COAP, ApriL 1 - March 31 GOM, COK, COTN L/ Wherever source material contains non-metric units, they have been converted using the above equivalents.  FOR OMCAL USR ONLY FREQUENTLY USED ABBREVIATIONS FOR OFFICERS ADA - Assistant Director of Agriculture AE - Assistant Engineer AEE - Assistant Executive Engineer CE - Chief Engineer CI - Canal Inspector DDA - Deputy Director of Agriculture DEE - Deputy Executive Engineer EE - Executive Engineer JDA - Joint Director of Agriculture LDC - Lower Division Clerk SE - Superintending Engineer VC - Vice Chancellor VEW - Village Extension Worker PRINCIPAL ABBREVIATIONS AND ACRONYMS AD - Department of Agriculture AED - Agricultural Engineering Department (Tamil Nadu) AP - Andhra Pradesh APAU - Andhra Pradesh Agricultural University APM - Adjustable Proportionate Module CAD - Command Area Development CADA - Command Area Development Authority CADD - Irrigation Utiiization and Command Area Development Department (see IU and CADD) CCA - Cultivable Commanded Area CCF - Construction Conversion Factor CWC - Central Water Commission DC - Distributary Committee DIC - District Irrigation Committee (Tamil Nadu and Andhra Pradesh) DOA - Department of Agriculture DOE - Department of Environment DSP - Dam Safety Panel FAO - Food and Agriculture Organization GCA - Gross Commanded Area GNP - Gross National Project GOAP - Government of Andhra Pradesh GOI - Government of India GOK - Government of Karnataka GOM - Government of Maharashtra GOTN - Government of Tamil Nadu IBRD - International Bank for Reconstruction and Development ICAR - Indian Council of Agricultural Research ICB - International Competitive Bidding ID - Irrigation Department or Irrigated Dry IDA - International Development Association IDB - Irrigation Development Board (AP) IIT - Indian Institute of Technology IMTI - Irrigation Management & Training Institute (Tamil Nadu) ISTF - Institutional Strengthening and Training Fund IU and CADD - Irrigation Utilization and Command Area Development Department (see CADD) iW - Irrigated Wet This document has a restricted distribution and may be used by recipients only in thc performance | nf -hri nffirial riltir Itc. rnnont I-v notI-- Ateri nh. rl-erncF- wihal InRnr^.h.s  LB - Left Bank LCB - Local Competitive Bidding LLC - Low-Level Canal M&E - Monitoring and Evaluation MC - Ainor Committee MITP - Maharashtra Induction Training Program MWR Ministry of Water Resources MSL - Mean Sea Level NA - Not Applicable NABARD - National Bank for Agriculture and Rural Development NARP - National Agricultural Research Project NSP - Nagarjunasagar Scheme NWMP - National Water Management Project NVP - Net Present Value OC - Outlet Committee O&M - Operation and Maintenance PAP - Project Affected Persons PCR - Project Completion Report PPAR - Project Performance Audit Report PPC - Project Preparation Cell PPF - Project Preparation Fund PPM - Project Preparation and Monitoring Wing (of ID) PPMC - Project Preparation and Monitoring Cell (Andhra Pradesh) PWD - Public Works Department R&B - Roads and bridges RB - Right Bank RD - Revenue Department RDS - Rajolibanda Diversion Scheme RWS - Rotational Water Supply SA - Special Account SCF - Standard Conversion Factor SDV - Supplementary Data Volume SMS - Subject Matter Specialist SRBC - Srisailam Right Branch Canal SRS - Sriramasagar Scheme SS - Sathanur Scheme TMC - Thousand Million Cubic Feet TN - State of Tamil Nadu T&V - Training and Visit (system of agricultural extension) UP - Uttar Pradesh VEW - Village Extension Worker VVS - Vanivilas Sagar Scheme, Karnataka WALAMTARI - Water and Land Management Training and Research Institute (Andhra Pradesh) WALMI - Water and Land Management Institute (Karnataka) WM - Water Management WMC - Water Management Cell WUG - Water Users' Group  GLOSSARY Anicut - Barrage Ayacut - Command Area served by a barrage Chak - Area served by watercourse, typically 15 ha to 40 ha District - The lowest administrative unit where the State Government is directly involved Hot Season - March to May Kharif - Wet Season (June to September) Laskar - Irrigation Department Field Worker Neergati - Man who irrigates on behalf of farmers Participating States - Andhra Pradesh, Karnataka, and Tamil Nadu Rabi - Dry season (October to February) Shejpali - Irrigation water supply system used in Gujarat and Maharashtra Taluka - A subdivision of a district Warabundi - NW India rotational water system PROJECT TERMINOLOGY Project - The entire National Water Management Project Subproject - The portion of the project carried out in each state Scheme - Irrigation entities, ise., Sathanur Scheme  INDIA NATIONAL WATER MANAGEMENT PROJECT Credit and Project Summary Borrower: India, acting by its President (COI) Beneficiaries: Government of India (COI) Government of Andhra Pradesh (COAP) Government of Karnataka (GOK) Government of Maharashtra (COM) Government of Tamil Nadu (GOTN) Amount: IDA Credit: SDR 93.2 M (US$114.0 M equivalent) Terms: Standard Re-Lending Terms: Prom the 0OI to the GOAP, GOK, 0OM, and GOTN as part of central assistance for State development projects on terms and conditions applicable at that time. 0OI would bear the foreign exchange risk. Project Description: The project will increase agricultural output by providing farmers with a more reliable and predict- able irrigation supply. To achieve this objective, it will strengthen the capacity, primarily in three participating States, to plan, implement and monitor improved O&M practices, and will support improvements in selected schemes in these states to demonstrate how such an approach can be effective. Provision is also made for pilot programs in other states. In addition, funds are provided to finance two other components related to the general objective of improved water management in India: (a) support for Induction Training of Engineers in Maharashtra; and, (b) an Institutional Strengthening and Training Fund to be managed by GOI id state agencies. The main element of the project will be implemented over a seven-year period; the Maharashtra Induction Training Program will be completed in five years; and the Institutional Strengthening and Training Fund would be fully expended in two years. The risks of the project stem mainly from the dif- ficulties of changing management patterns in existing irrigated areas and the shortage of experienced water management specialists.  INDIA NATIONAL WATER MANAGEMENT PROJECT STAFF APPRAISAL REPORT Table of Contents Page No. I. BACKGROUWND .......................................... 1 Agriculture and Irrigation in India .......................... 1 Bank Involvement in the Sector ............................... 2 II. PERFORMANCE AND POTENTIAL OF EXISTING IRRIGATION SYSTEMS ..... 4 Irrigation Management Systems in India ....................... 4 Recent Initiatives and Experiences ..........m.............. Past Experience as a Guideline for Future Planning 11 Rationale for Bank Involvement 12 III. THE PROJECT .............. ................................... .. 13 A. Project Origin, Scope and Objectives .....13 Project Origin ........... ,.................... 13 Project Objectives and Scope .......14 B. Detailed Features 14 Summary of Project Components ......................... ....... 14 Improved Water Management Components ......................... 15 Induction Training Program for Engineers in Maharashtra ...... 19 Institutional Strengthening and Training Fund (ISTF) ......... 21 C. Status of Preparation .....*......* ........................ 22 D. Ceneral .................................. 23 Dam Safety ................................................... 23 Environmental and Ecological Impact .......................... 23 IV. COST ESTIMATES, FINANCING AND DISBURSEMENT .................... 23 Project Cost Estimates ....................................... 23 Financing .... ........*... .... * * ... ....* * ..... .. ..... ... *... 24 Procurement ...... ............................................ 25 Disbursements . ..... ........**..... ........... 27 Accounts and Audits ............ .........*. ....... 28 This report is based on the findings of a Bank Mission which visited India in May 1986, and comprised of Messrs. B. Albinson, I. Naor, J. Berkoff, S. Singh, T. Bredero and Mr. H. Plusquellec (IDA); V. Branscheid, R. L. Chopra, and W. Barber (Consultanits). Word processing was done by Mrs. B. Scott with assistance from Mr. J. Maddix and Mrs. S. Douglas.  -ii- V. ORGANIZATION AND MANAGEMENT ....... .. .. .... . . . . . ....*.*. 29 General . ................ . . . 29 The GOI Water Management Cell (WMC) ....... 30 State-Level Organization . . ................. . . ..... *. . . . 30 Scheme-Level Organization ....................... . . . . 30 Selection and Approval of NWMP Subprojects .s..... .......... 31 Monitoring and Evaluation . . . . ...... . . 31 Training . ......32 Maharashtra Induction Training Program ............o....o.o... 33 Agricultural Supporting Services ............................. 33 VI. AGRICULTURAL PRODUCTION, FARM INCOMES AND COST RECOVERY ...... 34 A *AGRICULTURAL PRODUCTION s ......... es ... oo 34 Introduction ...... ...... ... . ............. .. ...... . 34 Impact on Agricultural Production in Selected Subprojects .... 35 - Rajolibanda Diversion Scheme ....... 35 - Vanivilas Sagar Scheme . .... 36 - Sathanur Scheme o. .. . . . .. .. . .... . . . ...... 36 Conclusions es ... .oso" . .606 37 B. FARM INCOMES *............. ****o ....*so ....... &... see* ... 38 Market Prospects and Prices 38 Farm Incomes . 38 C. OM FUNDING AND COST RECOVERY . ......... * ........ o... . ... *.s. 39 Background ..... 39 0OM Funding ...... ....... .*. 39 Cost Recovery .... . 41 Cost Recovery under the Project ............................ 43 VII. BENEFITS AND JUSTIFICATION ... ... ............... o... ..... 44 Introduction .. ....* .** *.* 44 Basic Assumptions for the Economic Analysis 44 Economic Analysis - Representative Schemes .... 45 Economic Analysis - Alternative Expenditures per Hectare sees. 45 Project Risks .... 47 VIII. AGREEMENTS TO BE REACHED AND RECOMMENDATIONS ................ 47 SCHEDULE A ...s.............. .... . 51  -iii- SUPPORTING TABLES Table 1 - Project Costs: Summary Accounts bv Year - 7 Year Program .............................. 52 Table 2 - Project Costs: Summary Account by Project Component - 7 Year Program .......................... 53 Table 3 - Andhra Pradesh: Detailed Cost Table ................. 54 Table 4 -Karnataka: Detailed Cost Table ..................... 55 Table 5 - Tamil Nadu: Detailed Cost Table .................... 56 Table 6 - Government of India: Detailed Cost Table ....0....... 57 Table 7 - Schemes Identified for Inclusion under Project ...... 58 Table 8 - Institutional Strengthening - PPC O&M Cell, Andhra Pradesh ............................ 59 Table 9 - Institutional Strengthening - PPC/O&M Cell, Karnataka 61 Table 10 - Institutional Strengthening - PPC/O&M Cell, Tamil Nadu 62 Table 11 - Institutional Strengthening - WMC ........ .......... 65 Table 12 - Staff Training Summary and Cost ..................... 66 Table 13 - Maharashtra Induction Training Program: Project Components and Costs by Year ............ o . .......... 67 Table 14 -Proposed Credit Allocation .......................... 68 Table 15 -Disbursement Schedule ............................... 69 Table 16 - Derivation of Economic Prices for Internationally Traded Crops and Inputs ............................. 70 Table 17 - Summary of Financial and Economic Prices ............ 71 Table 18 - Rajolibanda Diversion Scheme: Crop Yields, Areas and Production ........ ................. ....... 72 Table 19 - Vanivilas Sagar Scheme: Crop Yields, Areas and Production ............................ 73 Table 20 - Sathanur Scheme: Crop Yields, Areas and Production ............................ 75 Table 21 - Rajolibanda Diversion Scheme: Crop Yield & Input Data ....................... o76 Table 22 - Vanivilas Sagar Scheme: Crop Yield & Input Data .... 77 Table 23 - Sathanur Scheme: Crop Yield & Input Data ....o....... 78 Table 24 - Rajolibanda Diversion Scheme: Crop Budgets Per Ha . ..... o.&***# ...... 79 Table 25 - Vanivilas Sagar Scheme: Crop Budgets Per Ha ........ 80 Table 26 - Sathanur Scheme: Crop Budgets Per Ha ................ 81 Table 27 - Vanivilas Sagar Scheme: Coconut Crop Budgets ....... 82 Table 28 - Estimated Farm Budgets and Project Return at Full Development: Three Representative Schemes .......... 83 Table 29 -Summary of Economic Analysis ........................ 84 CHARTS Chart 1 - Organization Chart - Andhra Pradesh ................. 85 Chart 2 - Organization Chart - Karnataka ...................... 86 Chart 3 - Organization Chart - Tamil Nadu ..................... 87  -iv- ANNEXES IN MAIN REPORT Annex I - Assumptions for the Economic Analysis .......... . 88 Annex 2 - Selected Documents Available in the Project File .... 91 Annex 3 - Participating States in South India ......... 96 Annex 4 - Guidelines for the Preparation of a Scheme Proposal.. 103 MAPS IBRD 19790 - Andhra Pradesh: Irrigation Schemes IBRD 19799 - Karnataka: Irrigation Schemes IBRD 19777 - Tamil Nadu: Irrigation Schemes ANNEXES IN SUPPLEMENTARY DATA VOLUME I (SDV I) Annex 5 - The Rajolibanda Subproject Annex 6 - The Vanivilas Sagar Subproject Annex 7 - The Sathanur Subproject Annex 8 - Institutional Strengthening Annex 9 - Agricultural Research Annex 10 - Agricultural Extension Annex 11 - Linkages between Agriculture and Irrigation Agencies Annex 12 - Farmer Involvement Annex 13 - Monitoring and Evaluation Annex 14 - Training ANNEXES IN SUPPLEMENTARY DATA VOLUME II (SDV II) Annex 15 - Maharashtra Induction Training Program  IND'A NATIONAL WATER MANAGEMENT PROJECT Staff Appraisal Report I. BACKGROUND Agriculture and Irrigation in India 1.01 India has a population of about 800 M (mid-105), which is growing at an annual rate of about 2.2Z. Since 196G, per capita income has risen 1.2% per year, and in 1984/85 was close to Rs 3,420 (US$265). Although average incomes have increased, and access to public services has improved, about 25Z of the population continues to live in conditions of absolute poverty. Most of the poor are in rural areas and depend on agriculture for a living. 1.02 Agriculture continues to be the dominant sector of the Indian economy, contributing about 36% of GNP and engaging almost 70% of the total labor force. Foodgrain production accelerated substantially during the 1960s as a result of the green revolution and the introduction of high-yielding varieties of wheat and rice. During the Fifth (1974-79) and Sixth (1979-84) Plan periods, output has continued to increase by about 2.7% per year. About 60% of this increase can be attributed to greater yields, and the balance mainly to increased double cropping. Growth in wheat output (5.5% p.a.) reflects the outstanding performance of the Northwest States (Punjab, Haryana, Western UP). In contrast, rice production has increased by only 2.0% p.a. owing to the poor performnance of the Eastern States, and coarse grains have climbed only (1.5% p.a.) because of limited increases under rainfed conditions. However, the rising and more stable foodgrain output enabled India to withstand the adverse conditions created by monsoons in both 1979 and 1983, whereas serious disruptions were caused by only marginally more severe conditions in 1964 and 1966. The rising output of foodgrains has been accompanied by some diversification into higher-value cash crops (sugar- cane, cotton, etc.). At the same time, production of oil seeds and pulses has failed to keep up with demand. 1.(3 Since 1947, irrigation potential has risen from 19.5 M ha to 61.5 M ha (1983/84), with 93% of the potential estimated to be utilized. 1/ Sixth Plan achievements fell short of targets, but potential still rose at 2.25 M ha per year. This was the largest such program in the world and accounted 1/ Irrigation potential is created when construction to the Government outlet is complete, and is defined as the sum of the areas irrigated in the design cropping pattern. Potential is utilized when water is conveyed to the field. Not all States apply this definition in a consistent fashion.  -2- for 9.8X of plan outlays in comparison with a target of 11.6%. Ultimate potentias has been estimated at 113.5 M ha, 1/ or 60% of the cultivated area. 1.04 This expandion has conti.Ltu cf a steady growth in major and medium surface irrigation; rapid growth in (mainly private) groundwater developmen.t, which now accounts for about 402 of the irrigated area; and, more recently, some growth in minor surface irrigation. Development remains regionally concentrated, with four States (Uttar Pradesh, Bihar, Punjab, and Andhra Pradesh) accounting for more than 50% of the irrigated area. rhe exploited potential ranges from only 30% in Madhya Pradesh to more than 80% in Punjab ard Tamil Nadu. The groundwater (40 M ha) and surface potential (73 M ha) is expected to be fully developed early in the next century. 1.05 This increased irrigation accounts in large part for the improvement in agricultural performance. Irrigation has not only increased the area cropped and reduced crop losses in areas of inadequate rainfall, but it has also reduced the risks associated with rainfed farming, and thus encouraged diversified cropping, the use of HYV seeds, and the use of expensive inputs. Irrigated land, which constitutes only one-third of the cropped area, receives 85% of the fertilizer used. Bank Involvement in the Sector 2/ 1.06 Over the past 25 years, IDA has supported 43 irrigation projects through credits and loans amounting to $3,800 M. Of this amount, 61% went to the development of new schemes; 27% went to mcuernization or rehabilitation works; and the residual 12% was used to support schemes (for example, the replacement of barrages) that did not increase the irrigated area. Public groundwater development was a major component in four credits (Haryana II, Cr. 1319-IN, UP Public Tubewells I & II, Cr. 1004-IN, Cr. 1332-IN, and Bihar Public Tubewells Cr. 1737-IN). In addition, private groundwater exploitation has been extensively financed under agricultural credit programs (ARDC I, II and III, Cr. 540, 715, 947-IN) and the current NABARD (Ln. 2653-IN). Nine project PCRs and PPARs have been completed. 1.07 The basic thrust of Bank involvement in the sector has always been to increase the area irrigated, to ensure that the system is capable of serving the users, and to improve planning, design, and construction methods. However, the emphasis given to specific components of irrigation investments has varied. In the early 1970s, for example, attention was focused on the systems below the Government outlet--often up to 100 ha in size--which were supposed to be developed by the farmers. However, may of these systems lacked the necessary channels to distribute water to individual farms and plots. When experience with existing systems demonstrated that the standard 1! This excludes the potential of massive inter-basin eransfers from the Ganges-Brahmaputra to the South and West, estimated by the "National Perspective Plan for Water Resources Development" to be another 35 M ha. 2/ See Annex 3 for further details of Bank involvement and lessons learned in previous operations in the three participating states (Andhra Pradesh, Karnataka, and Tamil Nadu).  -3- design assumptions regarding seepage losses were grossly optimistic, lengthy discussions took place with GOI and the States concerned and considerable data were collected on this problem. The solution adopted in Bank-financed projects was to provide for the lining of all channels, with the proviso that lining would be omitted wherever analysis showed it to be uneconomical. In additicn, following unsuccessful attempts to encourage farmers to construct watercourses at their own expense, but with the support of institutional credit, OI made it national policy to have channels to turnouts serving areas of about 8 ha constructed at project expense. 1.08 Lending for irrigation has expanded very rapidly in the past 10 years and has supported several pioneering initiatives that have both enhanced the Bank's relationship with India and brought about institutional improvements within the country. For example, financing was extended to large "composite" projects, including three or four separately appraised major subprojects. Also financed in several States, were line-of-credit projects under which smaller subprojects were appraised according to criteria agreed by Project Preparation Cells in GOI and the States. At the same time, independent Dam Safety Review Panels, composed of Indian experts, were set up to oversee design and construction of dams. In separate initiative related to training of irrigation staff, Training Institutes in seven states were established for the express purpose of upgrading design and operation standards, especially of the minor systems. Research and Development components were included in some projects, which together with the Project Preparation Funds provided under two projects gave the Statesr GOI, and the Bank the means to address particular problems as they arose. In addition, a UNDP Project supported the use of foreign consultants to assist in establishing the WALMIs, to initiate pilot water management efforts, and to assist in other specialist tasks, including the production of guidelines for the design and operation of irrigation systems. At the project level, the most important technical innovation was the entirely new system of public tubewell irrigation intro- duced into UP. 1.09 Most recently, special attention has been given to two efforts in the irrigation sector. One was the preparation of projects on th.e Narmada River, which is the site of the largest irrigation scheme in India, and perhaps in the world. By becoming involved at the stage when fundamentai decisions were being made regarding water allocation among farmers, the physical structures required, the rules for system operation, the construction techniques for the project, and the handling of environmental and resettlement issues, the Bank was able to encourage GOI to improve all of these elements to standards previously unkrnown in India--and to consider institutional reforms, which are of at least equal importance. Second, special emphasis has been put on improving the operation of existing systems, which has been the primary concern in the formulation of this project.  -4- II. PERFORMANCE AND POTENTIAL OF EXISTING IRRIGATION SYSTEMS Irrigation Management Systems in India 2.01 Agroclimatic and socioeconomic conditions in India vary widely, and the irrigation systems that have evclved reflect this diversity, both in state policy and in individual schemes. Most surface systems can be classified into four main types: localization, warabundi, shejpali, and field-field. 2.02 Localization. In the southern states (Andhra Pradesh, Karnataka, and Tamil Nadu) where rice is generally the preferred crop, irrigation is tradi- tionally synonymous with paddy cultivation. "Localization" represents an attempt to limit the area under paddy and extend the irrigated area by desig- nating certain areas for "irrigated wet" (IW) crops such as paddy and sugar- cane, and others for "irrigated dry" (ID) or upland crops. Furthermore, to spread the benefits of irrigation as widely as possible, large areas were only partly served ("nonlocalized") even though they could be commanded by the system. The combined cropping pattern, whih was based on computed crop requirements and assumed operational efficiencies, was designed to coincide with the available water supply. To minimize capital costs and management problems, continuous supply was provided down to the Government outlet, the design duty for IW areas (1.80-1.20 1/s/ha, 40-60 acres per cusec) being about twice that for ID areas (0.72-0.60 1/s/ha, 100-120 acres per cusec). To make the best use of available resources, IW crops were in theory limited to low-lying areas with heavy soil, which naturally collected runoff, and ID crops were to be grown in higher areas, where soils were lighter. In practice, this was not always possible and other considerations frequently intervened (e.g., each village needed some area for subsistence paddy). 2.03 In theory, localization has merit, but it will not work unless ID farmers and nonlocalized farmers in the upper reaches are willing to abide by the localization pattern even though continuous and substantial flows for paddy are being passed through the system. 1/ Their unwillingness to do so has led to a breakdown of the localization distinction. The physical infrastructure is seen by the farmers as hindering irrigation in upstream ID and nonlocalized areas, and as inadequate in the lower reaches as upstream demand absorbs the available supply, a problem exacerbated by gross underestimation of seepage losses in the canal designs. In consequence, control structures are "modified" by the farmers, and no serious attempt has been made to formulate the rules and responsibilities. Instead, an ongoing struggle has developed between the farmers and the authorities, the scale of which makes it difficult to take action against offenders. 2.04 Furthermore, with abundant water during the early years of project operation, "temporary" conversions to IW were often allowed. Indeed, during the food crises of the late 1960s such conversions were encouraged, and it has subsequently been difficult to persuade farmers to revert back to ID crops. 1/ Even in ID areas, continuous flow to the Government outlet means that upper reach farmers along a distributary or minor, or within a water course com- mand, can support paddy at the expense of those lower down.  -5- Farmers have resisted such a change on the grounds that their fields were now levelled and prepared for paddy. (In some cases the whole local economy became geared to the conversion when a sugar mill was constructed.) Even where some farmers have welcomed a change to higher value, if more risky crops, the high water table induced by IW crops may discourage them from doing so. 2.05 The predominant pattern of irrigation in schemes designed for localiza- tion has thus been a concentration of paddy irrigation in the head reaches of canals; an intermediate or fringe area where farmers struggle to grow paddy or resort to ID crops in response to an unreliable supply; and tail reaches which are essentially planted to rainfed ID crops, with perhaps the occasional erratic irrigation, except where natural runoff allows for paddy cultivation. The spread of insecurity down the system is of considerable significance in relation to paddy. Rice is grown in stages, first in a nursery, which requires a lot of water over a small area, followed by transplanting into a puddled field, which requires a lot of water over a large area, and then by the growing period which requires continuous supply at a somewhat reduced rate. Where access to water is uncertain because the upper farmers control the supply, the timing of transplanting becomes uncertain. Since modern high-yielding, short duration rice varieties are very sensitive to the date of transplanting, risk-averse farmers instead plant long-duration varieties. The result is a patchwork of early planted, short duration varieties, and late planted, long duration crops. To serve the needs of both it is not uncommon that the Irriga- tion Department must run all the canals continuously for over seven months, leaving very little water for the rabi season. 2.06 Warabundi. The northwestern States of Punjab, Haryana, and parts of UP and Rajasthan operate the warabundi 1/ system. This system is based on a clear objective: to distribute available water as equitably and as reliably as possible. In principle water is allocated in proportion to land and the right to water goes with the land. The system was developed to meet the needs of water-short, river diversion projects, where the supply was both limited and erratic. It was clear that strong management would be needed to avoid the pressures from individual farmers trying to draw more than their share, and that the system should be transparently fair to users. It has survived intact as water has become more reliable and marginally more plentiful because of the construction of dams. 2.07 The strengths of the system are easily identified. The infrastructure is sized so that the capacity of successive channels is proportional to the area served, thus running a channel full will automatically provide the "proper" share of water to its command, and the sub-channels it serves will also automatically run full without adjustment of gates, or other interventions at lower levels in the system. The rules are also directly related to the management plan: irrigation schedules are drawn up and published, so that each farmer knows when water is expected to be provided; in addition priorities are 1/ Warabundi is in fact the system of operation below the Government outlet, while a separately defined rotational system governs operation above the outlet. However, the combined system is commonly referred to as warabundi. Warabundi requires rotational supply but the provision of a rotational supply does not ensure warabundi.  -6- assigned to sub-commands on a rotating basis, so that, if water is unexpectedly short, the order in which commands will be closed is known. Finally, within this framework, the responsibilities of the authorities and the farmers are defined. The Irrigation Department's job is relatively simple; they must ensure that the channels run full according to the schedules and priorities agreed. The ease with which this can be accomplished is, of course, strongly related to the fact that the infrastructure is specifically designed to suit the operational rules. The farmers' responsibilities and rights are also clearly defined; each farmer is entitled to draw the entire amount of water running in the watercourse in accordance with a schedule which divides the entire week into a series of turns, proportional in duration to the area of each holding. 2.08 Three other advantages of warabundi metit attention. First, each farmer has an incentive to use water efficiently because his share is limited, usually insufficient to irrigate his entire holdings, and thus has a high opportunity cost. Limited surface supplies also provide a powerful incentive for groundwater development which has become a very important factor in many areas. Second, the discipline set out in the rules of operation benefits all the farmers, creating a built-in robustness and stability to the system without the necessity for complex cooperation among the farmers. Third, the system's simplicity allows the farmers to monitor allocations. 2.09 While its strengths have ensured the success of warabundi, the system is not without its weaknesses. It is for instance not easy to accommodate crops with highly variable water requirements. In fresh groundwater areas, this can be offset by conjunctive use, but in saline areas cropping patterns may be constrained. Another problem relates to canal rotation practices in response to variable supplies. While in principle priorities are equitable and known, in practice operational problems relating to variable supply can lead to significant inequities between head and tail cormnands. 2.10 Shejpali. Cujarat and Maharashtra operate a demand system called shejpali, under which the farmers apply each season for permission to irrigate a particular cropping pattern. Applications are considered by committees at the project, command and distributary level. Sanctions are granted on the basis of water availability, channel capacities, and some attempt Vo avoid isolated patches of irrigation which are difficult to serve. Once sanction is granted, the farmer is authorized to draw water to irrigate the agreed cropping pattern, and the Irrigation Department is obliged to supply water. The infrastructure provided for operating the shejpali system is designed to allow much more flexible operation than is provided for in warabundi systems. Because the cropping pattern to be served is not known at the design stage, the channels and outlets are oversized for normal operation to allow for peaks. Many gates are provided to allow differential rates of supply to every outlet command (thus, in contrast to the warabundi system, low-level intervention into the running of the system is expected and provided for). The rules of opera- tion tend to define objectives rather than lay a basis for system operation; thus the Irrigation Department is responsible for supplying the "reasonable" needs of the sanctioned cropping pattern, while the farmers must not "waste" water, and must level their field "adequately". Operationally meaningful responsibilities are difficult to specify under these circumstances.  -7- 2.11 In an attempt to remedy these problems, a system called "rigid shejpali" is being introduced. Under this system, the sanctioned cropping pattern is made the basis for fixing schedules which are then to be strictly adhered to. Initial effects are positive: reliability improves, tailends are better served, and farmers are encouraged to grow higher value crops. The problem, however, is that the system design requires that the channels be run at varying levels depending on demand. This causes the discharges at the outlets to vary, making irrigation management difficult, and farmers often claim that their proper turns have not been fully provided. While the system is operating below full capacity, the slack can be used to "make up" local deficiencies. At peak demand, the system will be most difficult to operate and will have no slack, a combination that can create instability. 2.12 Field-Field. The fourth major system of irrigation in India is the traditional field-to-field irrigation practiced in the old, rice-intensive schemes, notably in the southern and eastern deltas. In complete contrast to the warabundi system, where intermittent flows are divided between farmers on the basis of time, field-to-field irrigation is based on continuous flows, divided into ever smaller streams, to the individual plot. Ad-hoc rotational supply is practiced for the land preparation phase. These systems have developed over many years and rely for their often considerable success on close, informal cooperation between farmers. Many of the modern barrage schemes have incorporated hundreds of age-old tank systems which complicate the management of the scheme. Few general remarks can be made about the infrastructure because it is largely the product of evolution rather than design. Similarly, the rules and responsibilities have developed to suit local circumstances. 2.13 Influence of the System on Irrigation Allocation. An important charac- teristic of the warabundi system is that it is a supply-based operation, under which the Irrigation Department follows clearly defined rules to distribute the available water as reliably and as equitably as possible. In this system, the farmer plants his crops and arranges his other agricultural activities to fit a pattern of water supply which, while not absolutely guaranteed, is nevertheless quite predictable. The general pattern of supply may be designed to match the general pattern of demand, for example, by giving consecutive weeks of flow during the wheat planting period, but the Irrigation Department is not obliged to attempt an exact match. 2.14 The other systems, especially shejpali and localization, are demand oriented, being based in theory at least on crop needs. Experience shows that it is often impossible to meet the complexities of individual farmer's demands with the infrastructure available. In times of high demand, the process, at best, becomes crisis management, reliability suffers, and most farmers even- tually adopt low-yield, risk-minimizing cropping patterns.  Recent Initiatives and E periences 2.15 Formal approaches to system management problems have been outlined in various GOI reviews and proposals: 1/ among such approaches are the pilot program in rotational water supply in Sriramasagar (Pochampad-Andhra Pradesh) and another in Mahi Kadana (Gujarat) which led to proposals for Bank-supported RWS projects in both Karnataka and Andhra Pradesh 2/, and the findings of the Commission for Irrigation Utilization in Andhra Pradesh. 3/ The pilot program in Sriramsagar initially hai a significant impact on agricultural production, 4/ but was located in areas with secure supplies to the Government outlet, and focused on farmer involvement and rotation below the outlet. The expectation was that farmers' organizations would create pressure to improve system manage- ment. However, the program had only a limited effect on the operation of the system as a whole, 5/ and when support for farmer organizations weakened, the program became ineffective. With the support of the Commission, however, the experience at Sriramasagar led the GOAP to adopt the "Integrated Water Manage- ment Program" under which canals are closed systematically one or two days a week to reduce water use and force water to tail-end areas (systematic canal operation), and detailed time-based rotation is introduced below the Government outlet (warabundi). This program has been widely adopted in AP schemes, and systematic canal operation in particular has had a significant impact. 6/ The introduction of warabundi (based on experience in Northwest India (para 2.06)) has been supported by a GOI centrally sponsored CAD program but has had mixed results, particularly in paddy areas (where it is difficult to maintain the channels below the outlet during the kharif seasor) and where main system supplies remain unreliable. 2.16 The RWS projects proposed for Andhra Pradesh and Karnataka were out- lined in a 1982 GOI proposal for a rotational water supply project (para 3.11). The objectives of all three proposals were welcomed by the Bank but their 1/ For instance; COI, "Manual on Irrigation Water Management," 1982. In the late 1970s, a GOI/CWC team visited schemes th-oughout India--including South India--and made detailed and specific recommendations for improving management of individual schemes. This initiative led to some improve- ments, but there was little systematic follow-up. 2/ IBRD/FAO Cooperative Program, "Andhra Pradesh Rotational Water Distribu- tion," Draft Preparation Report, 1981; and GOK, "Proposed Project on Better Irrigation Water Management through Rotational Water Supply," Preparation Report, 1980. 3/ GOAP, "Report of the Commission for Irrigation Utilization," 1982. 4/ See Annex 1, Draft Preparation Report for the AP Rotational Water Distribu- tion Project, op. cit. 5/ This is being tackled in the context of the recently approved AP Irrigation II project (Cr.1665-IN/Ln.2662-IN). 6/ See GOAP, "Review of the Impact of the Integrated Water Management Program at Nagarjunasagar Project," 1984.  -9- feasibility, phasing and institutional requirements were questioned. There- fore, as a first step, Bank staff were asked to pursue water management issues more actively in ongoing Bank-supported projects. In particular, they helped initiate programs in three widely differing schemes, two in South India (Nagar- junasagar in Andhra Pradesh and Upper Krishna in Karnataka) and one in Gujarat (Dantiwada): (a) Nagarjunasagar (900,000 ha) is a massive localized scheme for mixed paddy and ID crops (para 2.02). Building on the earlier Integrated Water Management Program (para 2.15), GOI in 1984/85 introduced an approach--which was expanded in 1985/86--that added a number of fea- tures, particularly in the area of reservoir operations (carryover for a guaranteed early start to the following kharif season), water budget- ing, intermittent supply for ID crops, and the timing and completion of cropping activities. Despite initial teething problems, the program has been successful and is to be expanded further in 1986/87. The area under irrigation expanded significantly faster than water use (in 1985/86, the irrigated area was 68% higher than in 1983/84, whereas water use rose only 17.5%) 1/; there was a major shift to ID crops (from 30% of the area irrigated to 46%); yields of paddy in the 360,000 ha of the early sown water management zone rose markedly above past levels (4.75-5.5 t/ha compared to 3.5- 4.5 t/ha); 2/ and there was a large increase in the area of pulses grown on residual moisture (from about 20% to 37% of the paddy area). Unfortunately, a serious pest attack (which spread throughout the region) reduced the yields of cotton, the major ID crop; however, this reduction cannot be attributed to the water management program. Although other factors may have contributed to some of the favorable results (e.g., changes in the relative profitability of crops) and a partial switch back to paddy has occurred this year, in general, the program achieved its major objectives. (b) Upper Krishna is a major new project still in the early stages of development. In operating plan initially proposed, rotation of irrigated areas with the command was such that in any one year a third of the area would be under kharif crops, a third under rabi, and a third under two seasonal crops, and these areas were to rotate year by year. Although such an approach might improve equity, in comparison with "localization" based on fixed areas, it would still have most of the drawbacks of traditional practice (para 2.02). Hence, an alterna- tive system was adopted, which was similar in many respects to the Northwest system (para 2.06) and is based on the allocation of pre-determined supplies in proportion to the irrigable area. Dis- tributary canals are run full or not at all; an irrigation schedule is followed--it is announced clearly in advance and consistent with general crop needs; warabundi schedules are introduced below the out- let; and ungated distribution systems are being constructed. Although 1/ Large areas on the Left Bank were brought under irrigation for the first time and benefited from the improved operations. 2/ Lowest and highest average yield oy Taluka.  -10- it is impossible to compare what would have occurred with the original proposals, the schedules have been generally followed and farmer response has been favorable. (c) Dantiwada Irrigation Scheme (45,000 ha) is being modernized under the Gujarat Medium II Project (Cr.1496-IN). 1/ As elsewhere in Gujarat and Maharashtra, it was operated according to the shejpali system of crop sanctions under which farmers apply for irrigation on a seasonal, annual or longer-term basis (para 2.10). To counter the problems associated with this system, a warabundi program was initiated in 1984 under which all farmers receive a share based on area; delivery schedules are predetermined (allowing in rabi for carryover); outlet capacities are adjusted in proportion to the area served; and warabundi schedules are introduced below the outlet. While drought severely curtailed rabi cultivation during 1985/86, evidence from the previous year suggested a 29% increasc in cropping over the control area com- bined with a 15% savings in water use. Significant yield benefits were also claimed. Given these results, the program has been continued and guidelines based on this experience have been issued by COG to assist operations in other schemes in the state. 2.17 Initial experience with these three programs and two others (Hula in Maharashtra and the Rajasthan Canal) was reviewed in a report issued in October 1984, and in more recent reports on Nagarjunasagar and Dantiwada. 2/ In addi- tion to these programs, many other initiatives have been taken both in Bank-supported and other programs. These include programs in shejpali systems, e.g., in Mula, where emphasis has been given to mobilizing farmer involvement, in Jayakwadi, where long experience with the pilot and farmers' projects has led to programs supported under the Maharashtra Water Utilization (Cr.1383-IN) and Maharashtra III (Cr.1621-IN) Projects; and in Mahi where promising results have been obtained from a USAID-supported operational research program; in northwest warabundi systems, e.g., pilot programs in Haryana (Cr.1319-IN) and Rajasthan (IFAD Ln. 32-IN) which have shown the agricultural impact of provid- ing farmers with predetermined fixed schedules, and overall scheduling models for Haryana and Punjab which show promise for making the system as a whole more responsive to local requirements); 3/ and in southern systems, e.g., in Periyar Vaigai where programs have shown promise for improving scheduling of supplies and integrating tanks, and in the Cauvery Delta and Thambaraparani where reduced supplies and rotational issues had a significant impact on yields in 1985/86. 1/ Modernization started under Gujarat Medium Irrigation I project (Cr.808-IN) which closed in 1984. 2/ World Bank, New Delhi Office, "Nagarjunasagar Project--Evaluation of the 1985/86 Water Management Program," May 1986, and V.B. Patel, "Review of Dantiwada Water Management Program," May 1986. 3/ See Dr. S.N. Narayanamurthy, "Methods for Improving Canal Regulation in Northwest India," December 1985.  -11- Past Experience as a Guideline for Future Planning 2.18 Several important conclusions can be drawn from these experiences and from more general experience from schemes in India. First, it is not possible to provide service to individual farmers and plots with the infrastructure available in most surface irrigation schemes in India. Consequently, at some level in the system, control must be transferred to the farmers. Second, farmers cannot organize to perform their role in the distribution of water if the supplies they receive are unreliable. Third, reliable supplies can only be assured if the operational plan is realistically framed in relation to the available infrastructure. Ideally, each farmer would receive his allotted share of water on a predetermined schedule, allowing him to plan his operations with minimum uncertainty. In many schemes, this standard is a long-term objec- tive, which can be reached over t:me through successive rounds of infrastruc- tural and operational improvements. 2.19 Deciding on how to intervene in any particular case is neither simple nor easy given the large number of parameters that describe a given irrigation system and its operations. Any interventions must reflect an understanding of how the system presently operates and why it has evolved as it has. They must be responsive to the established socio-economic environment and to evolving agricultural conditions, and be realistically framed in relation to the avail- able infrastructure. They must also be consistent with long-term objectives for the scheme and the river basin concerned. 2.20 A fundamental consideration relates to the water supply. Not only may it differ from design expectations but it changes in relation to the irrigated area as the scheme and river basin develop and can vary enormously from year to year. Furthermore, canal supplies are only a part of the total water avail- able. Operating practice has often evolved to suit the convenience of canal management rather than to make best use of all the resources available (rain- fall, groundwater, tanks, and re-use systems, etc., as well as canal supplies). A re-evaluation of the water resource is therefore an essential first step, and, taken in association with evolving agricultural requirements, may suggest changes in the rules governing reservoir operations, in the ways that water rights are satisfied or in water distribution practices. While groundwater development would not be specifically funded under the project, conjunctive use by privately developed shallow wells does provide flexibility and improves the efficiency of many projects. 2.21 The modifications to be introduced must respond to the dominant crop mix, how this is changing and the extent to which operational practice can encourage further desirable changes. An agricultural plan consistent with the available water resource must therefore be prepared with an operational plan for water delivery to meet the revised agricultural objectives. Where rice predominates, the priority needs are usually to support early nursery estab- lishment (in order to encourage short duration HYVs, maximize use of rainfall and ensure early completion of the kharif crop) and to coordinate transplanting on a block-by-block basis (so that water is not wasted in the process of serv- ing scattered areas of late-planted crops and to allow an even and early start to the rabi crop). For upland crops, the priority may also be to guarantee the timing of the kharif crop, while during rabi a supply based system, such as that practiced in the Northwest (para 2.06), has much to offer since it allows  -12- the responsibilities of the Irrigation Department to be clearly and realisti- cally defined. The level of the system to which the scheduling is fixed must be carefully related to the infrastructure available since, if the plan is over ambitious in relation to this infrastructure, the prospects for involving the farmers in further improvement will be limited. Among the most problematic schemes are those with mixed paddy and upland crops. Often under these condi- tions, interventions will seek to limit paddy to low-lying areas through supply-based scheduling practices coupled with special paddy issues, for example, to tanks. In other cases, this may not be possible and more complex operations may be inevitable. A common situation in high rainfall areas is a paddy/upland crop rotation where the problem is to distribute limited rabi supplies through an irrigation system designed for paddy (which consequently may have few control structures, limited field channels and higher channel capacities than needed for upland crops). 2.22 The operational plan defines the water delivery service to be provided to the farmers. It must be acceptable to them, and can only be finalized following discussions with them and the approval of the scheme's managing board or committee. To ensure its implementation, responsibilities have to be assigned, other supporting services, input supply, extension, etc., have to be programmed to meet its requirements, and farmers must be fully informed of what to expect. Available financial resources need to be expanded in support of the operational plan. If these are restricted to normal O&M allocations, the plan will have to be related to the existing infrastructure, maintained as far as possible to meet its requirements. Where additional resources are provided (as proposed under the NWMP), the plan can be modified, supported by enhanced 0&M allocations and scheme investments designed to upgrade the system for better management. If substantial additional resources are available, a major modern- ization of the system may be desirable, possibly accompanied by a radical change in the operational plan. Whatever interventions are proposed, satisfac- tory arrangements must be made for monitoring and evaluation so that the plan can be modified in subsequent years to reflect the successes and failures of prior experience. 2.23 Each scheme is different, and the operational interventions that are appropriate will vary widely. Most non-rice areas will adopt a rotational regime. In rice areas, rotation is commonly used in the land preparation phase, which requires very high rates of application, followed by continuous flow in the growing phase. Nevertheless, the above suggests that there is sufficient practical evidence to support a systematic approach to operational planning, and that the process is similar for all schemes. Guidelines for preparing such a plan are included in Annex 4, and the proposed project would seek to initiate such an approach in selected states and schemes. Annexes 5, 6, and 7 of SDVI summarize proposals for three demonstration schemes to illustrate the application of appropriate technology. Rationale for Bank involvement 2.24 Improving the productivity of existing irrigation investments is a primary sectoral objective of both the Bank and GOI. Although irrigation has contributed greatly to India's success in increasing foodgrain production over the past 20 years, the returns to the investment made have been lower than they would have been if water management had been given adequate attention.  -13- 2.25 Improved management of irrigation projects depends fundamentally on initiatives by the field staff who plan, direct, and execute canal system operations. But outside agencies, such as the Bank or the central govern- ment, can have an important and sometimes essential catalytic role in, the process. 1/ First, the process of defining the plan of operation promotes constructive questioning of traditional norms and practices; second, it provides a forum where new ideas--often provided by the existing staff--can legitimately be discussed with senior officers; third, the outside agency can often bring proven ideas from other states, or identify approaches that have already failed elsewhete; finally, by formalizing these activities into a project, with institutional linkages to the state and central governments, the importance and benefits of improved canal operation. III. THE PROJECT A. Project Origin, Scope and Objectives Project Origin 3.01 GOI first approached the Bank with a proposal for a national water management project in 1982. 2/ Its objectives were supported by the Bank but questions were raised concerning its feasibility, phasing and institutional requirements. As a first step, therefore, Bank staff were assigned to pursue water management issues more actively under ongoing Bank-supported projects and, in particular, help initiate the three programs reviewed in Chapter II (para 2.16). Given the encouraging results, and the other evidence reviewed above, GOI in late 1984 established a Water Management Cell (WMC) in the Ministry of Irrigation (now Water Resources), and the Bank and GOI together agreed to explore the potential for a water management project in selected States. Guidelines for preparing scheme proposals were developed, 3/ and a joint GOI/Bank team helped prepare three "model" proposals. 3.02 These proposals met with a favorable State response, and in April 1985, GOI decided to initiate detailed preparation in three states (Andhra Pradesh, Karnataka and Tamil Nadu). 4/ If the concept was successful, other states were to be included at a subsequent stage. Assisted by the joint GOI/Bank team, each of these States prepared an overall project report, finalized a detailed scheme proposal and began preparation for a number of other schemes (see para 3.32). 1/ See in particular the extensive appraisal reports for the Narmada program in Gujarat. 2/ GOI: "National Program on Rotational Water Supply System" (Draft for Discussions with the World Bank), 11 April 1982. 3/ See Annex 4: "Guidelines for Preparing a Scheme Proposal". 4/ See Annex 3 for background information on these three States.  -14- Project Objectives and Scope 3.03 Objectives. The purpose of the project is to increase productivity and farm incomes in existing irrigation schemes through a more reliable, predictable, and equitable irrigation service. To achieve these objectives and facilitate extensien of similar activities throughout the country, the project would help participating States and OI develop an institutional capacity to plan, implement, and monitor improved O&M practices, and would provide for selected schemes for low-cost infrastructural improvements designed to support an improved operational plan. In each case, the most appropriate technique of water supply would be adopted. 3.04 Scope. Sixteen schemes have been identified by the participating states. They command about 1.68 M ha, of which 0.46 M ha would be included in the improvement programs to be financed under the project. Provision would also be made for a further 0.12 M ha in schemes yet to be identified; these would be divided about equally between the participating States and pilot programs in other States. Thus, the project is expected to cover about 0.58 M ha directly (Table 7). 3.05 Scheme Formulation. The most important element of scheme formulation is the preparation of an operational plan. On the basis of water availability, system characteristics, and agricultural options, the plan defines how the system would be operated with respect to the timings and quanUities of water deliveries and the responsibilities of those involved (paras. 2.18-2.23). The basic elements of and procedures for formulating an operational plan are presented in Annex 4. Required staffing, budget, and facilities for O&M would form part of the plan, which would also program the scheme investments to be supported by the project. B. Detailed Features Summary of Project Components 3.06 The principal components of the main part of the project (improved water management) would be: (a) Scheme Investments required to implement the operational plan. These would include repairs and renovations to existing structures; provi- sion of new control structures; field channels and drains; measuring devices; and investments in upgraded O&M facilities and equipment, including housing for staff and transport and communications equip- ment, as required to execute the operational plan. (b) Engineering and Administration required to supervise the effective implementation of the scheme investments. (c) Incremental Recurrent O&M Expenditures to support the improved opera- tional plan and maintain it after the completion of the scheme investments. (d) Institutional Strengthening of the State Irrigation Departments, with particular emphasis on helping them develop O&M planning and support units in the states. In addition, the project will strengthen the  -15- Water Management Cell (WMC) in the Ministry of Water Resources (MWR) so that it can provide policy direction, act as a catalyst for extending the approach across state boundaries, and administer and coordinate the NWMP as a whole. In each case, the project would support staffing costs, buildings, equipment, vehicles, and associated recurrent costs. (e) Training Programs. At the scheme level, an orientation training schedule for field staff and farmers would be defined as part of the scheme proposal; the project would also provide support for technical and orientation training programs for scheme staff, State O&M staff, and staff of the WMC. Study tours within India and limited overseas visits would be provided for, along with support for the infrastruc- tural development of the Irrigation Management and Training Institute in Tamil Nadu. (f) Special Studies and Programs. Provision would be made for consultancy services and for other assistance designed to strengthen specific aspects of water management (e.g. monitoring and evaluation, computer applications, communications, dam safety, etc.). 3.07 The project would support two other components related to the general objectives of improved water management in India: (a) The Maharashtra Induction Training Program, which will support ini- tial vocational training for engineers recruited to the Maharashtra Irrigation Department; and (b) An Institutional Strengthening and Training Fund (ISTF) to be managed and implemented by GOI and state agencies along the lines of two previous Project Preparation Ft'nds. Improved Water Management Components 3.08 Scheme Investments. The physical investments proposed for the scheme improvement component of the NWMP (about 80% of the cost) would cover deferred maintenance, corrective additions required to implement the opera- tional plan, and enhanced O&M facilities and equipment. On the basis of the cost estimates for 10 schemes involving some 390,000 ha (62.5% of the total area to be covered under the project), the upper limit for scheme investments has been set at Rs 2,500/ha. This figure is low enough to ensure that the approach can be replicated and that attention will be concentrated on the highest-priority investments. In cases where the State decides to invest in more intensive rehabilitation, project disbursements would be limited to investments of up to Rs 2,500/ha at 1986 prices, including physical contin- gencies. The composition of scheme investments is illustrated in the three schemes outlined in the following table:  -16- Rajolibanda Vanivilas Sagar Sathanur Diversion Scheme (VVS) Scheme (SS) Scheme (RDS) (Andhra Pradesh) (Karnataka) (Tamil Nadu) Rs M (1986 prices) Civil Works 69.205 30.660 35.290 Measuring Devices 0.910 0.450 1.157 Telephones/Communications 0.340 0.200 0.300 Operators Housing 1.300 0.800 1.005 Rain Gauges/Current Meters 0.100 0.050 0.045 Tools and Plant 0.400 0.100 0.900 Transport (jeeps, motorcycles, bicycles) 0.620 0.300 0.600 Office Buildings - 0.500 - Office Equipment 0.100 0.050 - Total 72.975 33.110 39.297 Base Cost Rs/ha 2060 2365 2158 3.09 Incremental O&M. A detailed O&M plan will be prepared for each scheme and included in the scheme proposal. This plan will specify the actual O&M funds required year by year to manage the schemes effectively and efficiently. A clear distinction will be made between: (a) recurrent expenditures required to implement the operational plan, principally staff at both the engineer (officer) and nonprofessional (work supervisor, laskar, support staff) levels and associated recur- rent expenditures (such as, transport and utilities). To the extent consistent with the requirements of the operational plan, staff increases will be kept to a minimum--for example, through reassign- ment or retraining of existing personnel--so that the past erosion in the share of maintenance in the total will be reversed; and (b) maintenance expenditures required to maintain the physical infrastructure. Agreement has been reached with the GOAP, GOK, and GOTN that they would, for all NWMP schemes, provide the O&M requirements as defined in the scheme reports and which would be satisfactory to GOI and IDA (paras 5.08-5.09). 3.10 Provision has been made under the project for scheme specific incre- ments in the O&M amount in excess of the average of the three years prior to the first year in which O&M support is provided. Typical analyses are given in Annexes 5, 6, and 7 in SDV I, which can be summarized as follows: Subproject Current Pro osed Incremental ___-----------Rs/ha --------- Rajolibanda Diversion 50 146 96 Vanivilas Sagar 90 170 80 Sathanur 50 100 50  -17- These requirements (expressed in 1986 prices) and analyses of other schemes indicate that incremental O&M would be about Rs 80/ha. This figure has been included in the cost estimates, phased as follows: Rs 40/ha in the first year in which incremental expenditures are provided, Rs 60/ha in the second year, and Rs 80/ha in the third year and subsequent years. The year in which incremental O&M will start will be defined in the scheme report; for purposes of estimation this has been assLmed to be the third year of each scheme programe 3.11 Institutional Strengthening. Project Preparation Cells (PPCs) will prepare schemes with the assistance of scheme staff and other State agencies. The specific organizational arrangements will vary from State to State. In each case, the incremental costs related to planning and management arrange- ments would be financed under the project, and would include: (a) staffing (engineers, surveyors, draftsmen, supporting staff, etc); (b) materials and equipment; (c) technical assistance from other agencies (e.g. groundwater); and (d) associated recurrent costs. Not only will the PPC itself be strengthend, but arrangements have been made to strengthen the units undertaking detailed survey and design work. In Andhra Pradesh, subdivisions have been establi! ed under the State Central Design Organization on secondment to the PPC; in Karnataka, staff are to be reallocated within the existing regional organizations (which are generally responsible for most survey and design work); and in Tamil Nadu, subdivisions have been established under the PPC itself. In addition to preparing schemes, the PPCs would manage the implementation of the NWMP. They would monitor project progress, report to GOI and the Bank, and maintain project accounts. A Project Accounting Unit (PAU) would be established in each PPC to handle these project accounts, as well as requests for financial provi- sions and collation of expenditures for reimbursement from IDA. 3.12 State O&M Cells. The project would strengthen the capacity of the State Irrigation Departments to provide OM policy, planning, and implementa- tion support. In each participating state, an O&M cell would be established to assist in: (a) the overall direction of State O&M activities; and (b) the planning, implementation, and monitoring ot scheme-level OEM programs. The relationship between these units and the PPCs would differ from State to State and evolve over time (para 5.04). Their responsibilities and the specific arrangements in each State are described in SDV I, Annex 8. 3.13 Charts 1-3 show how the PPCs and O&M Units would be placed in the various State organizations. Tables 8, 9, and 10 show their estimated staff- ing levels and costs.  -18- 3.14 The GOI Water Management Cell (WMC) in the MWR was established in late 1984 to help prepare and promote the NWMP. Its technical staff (cur- rently comprising one Chief Engineer (WM) and one Joint Commissioner (WM)) would be expanded under the the project, and would be provided with appropriate support staff, transport, office and dra'lting equipment. Table 11 shows the estimated costs of the WMC. 3.15 Training. Training would be an integral part of the project and crucial to its success. It would be provided at five levels: (a) Orientation Training for senior GOI and State officials in the over- all objectives, requirements, and policy implications of the programs supported under the project. (b) Training in Planning, particularly in the preparation of scheme operational plans, to be given to staff of the O&M units and selected senior scheme-level officials. (c) Training in Design for the engineers responsible for the revised hydraulic designs needed to support the operational plans implemented under the project. (d) Implementation and Operation Training to ensure that the scheme staff fully understand what is proposed under the project. Such training would be given to: (i) scheme-level officers (here the training would focus on the operational plan proposed); and (ii) scheme-level field workers, to ensure they understand what their responsibilities are and to help create a collegiate approach to system operations. (e) Farmer Training to ensure that farmers understand and agree with the system proposed and are aware of their responsibilities and those of the Irrigation Department. Training programs for the demonstration schemes with respect to (d) and (e) have been prepared in the scheme proposals. Additional details are presented in para 5.14 and SDVI, Annex 14. 3.16 Provision would also be made to assist in the development of the Irrigation Management and Training Institute in Tamil Nadu. Detailed architectural designs have been prepared, and Phase I of the Institute is being established with assistance from USAID. 1/ The project would provide funds to complete the construction of Phases II and III of the development programn. Details ar3 provided in SDVI, Annex 14. AP has a similar institute, which was established under the AP Composite I Project 1/ The Institute was to have been included in a proposed Cauvery rehabilita- tion project, but this has been delayed. USAID made limited funds avail- able to allow an early start.  -19- (Ln.1251-IN) for the Water and Land Managemeiut Training and Research Institute (WALMTARI). In Karnataka, the Water and Land Management Institute (WALMI) already conducts training courses in temporary facilities. Permanent structures are planned under the Karnataka Irrigation II Project. These three institutes would play an important role in providing training for the NWMP (para 5.14). 3.17 Special Studies and Programs. Limited provision is made under the project for consultancy services and other assistance designed to strengthen specific aspects of water management and related matters. The funds would be allocated to the participating States (Andhra Pradesh, Karnataka, and Tamil Nadu) and GOI, and would be managed by the respective PPCs and the WMC of GOI. These funds would cover the following items: (a) evaluation studies, by independent agencies, of the impact of opera- tional plans in specific schemes (para 5.12 and SDVI, Annex 13); (b) consultancy services for developing computer applications in system operations; (c) consultancy services and remote sensing equipment for developing appropriate communication systems for system operations; and (d) the costs associated with field visits by the Dam Safety Panels (para 3.35). Induction Training Program for Engineers in Maharashtra (MITP) 1/ 3.18 Background. Irrigation Departments in the states recruit virtually all their professional engineering staff from young civil engineering graduates, whose education covered general civil engineering and included little, if any, irrigation technology. Modern irrigation has developed into a complex multidisciplinary technology involving engineering, hydrology, agronomy, economics, sociology, and management sciences. Few institutions are equipped to teach comprehensive courses in all these disciplines. 3.19 The objective of the Induction Training Program would be to impart an adequate irrigation engineering background to all graduate civil engineering recruits of the Irrigation Departments before the end of their probation period. The engineering program would include advanced academic training in irrigation engineering disciplines and practice-oriented training in the field on-farm development works, irrigation practices, irrigation agronomy, and related subjects. The program in Maharashtra would be the first of its kind in India and, if successful, is expected to be the forerunner of similar programs in other States. 1/ For a full description, see SDV II, Annex 15.  -20- 3.20 The Training Program. The Irrigation Department of the Government of Maharashtra recruits about 60 civil engineering graduates every year. The recruits would be divided into two groups. The courses would be essentially the same for both groups, but the emphasis would be slightly different for group (A) Operation and Management Engineers who will operate, maintain, and manage the systems; and group (B) Design and Construction Engineers who will plan, design, and build projects. Engineers in group (A) would also be trained to advise farmers on on-farm works and irrigation practices. Each group would include about 30 officers. 3.21 The program would include Lhree main types of training, each which would be carried out in a different training institution, where most of the necessary infrastructure exists (the project-related costs to be funded are summarized in Table 13: (a) Basic Training will ensure that the recruits understand governmental responsibilities and work procedures and the interrelationships between the various government departments engaged in irrigation development. This three-month training is already given to recruits at the Engineering Staff College at Nasik, and would continue basi- cally unchanged in the future. (b) Technical Training would teach modern irrigation technologies and practices, particularly in areas of project planning, the operation of irrigation systems, on-farm irrigation works, and general water use management. The course would be divided into academic and prac- tice-oriented periods of about four months each, and would be given in two Institutions: (i) Academic Training at the Indian Institute of Technology (IIT), Bombay. This one-semester academic course would fill gaps in classical subjects such as mathematics, statistics, hydraulics, open channel flow, hydrology, operations research, and computer science. At present, most of these subjects are not adequately covered in the undergraduate curriculum of the gEneral civil engineering degree. (ii) Practice-Oriented Training at the Water and Land Management Institute (WALMI) in Aurangabad. Recruits would spend one semester at the Institute for training in the operation of irrigation projects, water scheduling (which would cover the techniques required to supply water to the farm gates reliably and at the correct time) and subjects related to irrigated agriculture, including the preparation of land for irrigation, on-farm .rrigation practices, the supply of agricultural inputs, and efficient water and land use. 3.22 A number of specially designed courses to be given both at IIT and at WALMI which will relate to specific problems of irrigation in Maharashtra (black cotton soils, difficult topography, etc.).  -21- 3.23 The training program would be organized so that only one group of trainees (about 30 participants) would receive instruction (at the IIT or WALMI) at any one time. Thus, the infrastructural facilities and faculties in the two Institutions would be utilized throughout the year. 3.24 MITP Subcomponents. The MITP subcomponents include: (i) preparation of course materials; (ii) provision of modern audio/visual, computer1 and other teaching aids; (iii) incremental faculty staff costs; (iv) incremental accommodation and classrooms; (v) direct course costs; (vi) consultancy services; (vii) training of faculty; and (viii) incremental operation and maintenance costs. 3.25 The total estimated cost would be about Rs 40 M (US$3.1 M), of which about US$1.2 M would be in foreign exchenge (Table 13). Institutional Strengthening and Training Fund (ISTP) 3.26 A Project Preparation Fund (PPF) was established under the Second Cujarat Irrigation Project (Cr.1011-IN, in 1980, US$175.0 M) to encourage the States to undertake more detailed data collection and project preparation, and to promote the introduction of new planning techniques. This credit provided US$10.0 M over a two-year period for the preparation of irrigation projects for Bank Group Financing. Continuation funding of US$10.0 M has been provided under the Uttar Pradesh Tubewell II Project, (UP II, Cr. 1332-IN, 1983, US$91 M). 3.27 With support from the PPF, planning agencies for the Narmada Basin Development have been created in the States of Gujarat and Madhya Pradesh. In addition, project preparation and monitoring (PPM) cells for five States have been established in the Central Water Commission (CWC) to assist the States in preparing medium-sized irrigation schemes. PPM cells have also been set up in GOI's Ministry of Water Resources for the coordination and support of selected planning studies for Narmada Basin Development. Exten- sive use has been made of Indian consultants. Foreign consultants, in joint ventures with local firms, have studied navigation on the Narmada River and have carried out a comprehensive drainage investigation. Furthermore, a panel of Indian and foreign experts has advised State organizations on the design of several major dams. Funds have also been provided for equipment and pilot operations. Training programs have been supported both in India and overseas. 3.28 The PPF has been instrumental in gradually strengthening the States' ability to plan and prepare irrigation and multipurpose water resources projects. It has also helped to build up the domestic consulting industry and to introduce modern planning techniques such as systems analysis. In short, the PPF has made an important contribution to the modernization of India's irrigation sector. The US$10.0 M provided under the existing is expected to be fully disbursed by December 1986. 3.29 The institution-building efforts started under the first two PPFs would be consolidated through the provision of an additional US$10.0 M for continuing the improvement in the preparation of irrigation projects in India. The present approach of tailoring the support to the individual requirements of the States, for programs mutually agreed on by the Bank and  -22- GOI would be maintained and, thus, the expected expenditures cannot be broken down in detail. To reflect the proposed utilization of the fund, its name has been changed to Institution Strengthening and Training Fund (ISTF). lt is expected to be used for the following purposes: (i) support for the preparation of irrigation projects, primarily for studies, data collection, other project preparation activities, and pilot works; (ii) continued support of existing and new PPM cells and strengthening of project preparation and monitoring activities at MWR--such support would include the purchase of equipment (such as calculators, microcomputers, and office and other techni- cal equipment, including survey instruments), vehicles, and other nonrecur- ring expenditures; and (iii) technical assistqnce (including consultant services, training, and study trips for specified components under Bank-assisted projects in India). ISTF would be managed by MWR. Activities to be financed would be presented in a format mutually agreed by the Bank and GOI. The authority to approve such activity would rest with the Joint Secretary, MWR. Agreements have been reached with COI that it would: (i) cause each state receiving ISTF proceeds to utilize the funds allocated in accordance with procedures satisfactory to IDA; and (ii) submit an annual progress report not later than June 30 each year, commencing June 30, 1988, itemizing and describing progress and the expenditures under the ISTF incurred in the previous year ending March 31. C. Status of Preparation 3.30 The schemes that the participating states would like to see included in the project are given in Table 7 together with their cost estimates. In addition, the table gives provisional cost estimates for schemes yet to be identified. 3.31 Programs for completing the necessary surveys, designs, and estimates for the physical works to be undertaken in each of the appraised schemes have been agreed by IDA. Modified control structures (self-regulating weirs and baffle-type offtakes) are to be introduced where appropriate, and detailed designs are being prepared. A condition of negotiation was that GOAP, COK, and GOTN finalize the detailed designs and estimates for the first year of the construction program of the proposed project. 3.32 In addition to the three appraised subprojects, draft reports have been prepared for Kurnool-Cuddapah Canal and the Nagarjunasagar left and right bank schemes in Andhra Pradesh, Bhadra, and Tungabhadra; the left bank schemes in Karnataka and Kodayer; and the Thambaraparani schemes in Tamil Nadu. These schemes cover about 325,000 ha, and together with the 67,500 ha included under the appraised subprojects account for about 63Z of the area to be covered under the project. Two other proposed subprojects (Mylavaram and Thandava schemes in Andhra Pradesh, covering about 50,000 ha) are also in preparation. 3.33 The physical works for individual subprojects would normally be completed within three years, but may extend into a fourth year for rela- tively large schemes. The number of subprojects undertaken at any one time will depend on the overall plan allocations and the annual budgets provided. The overall seven-year implementation progrr.n that has been prepared is consistent with State allocations under the ongoing Seventh Five-Year Plan and with the Bank's standard disbursement profile for irrigation projects in  -23- India. If additional funds can be provided in the early years, project implementation may be completed in less than seven years. No new subproject would be started after June 1, 1990. 3.34 The Maharashtra Induction Training Program (NITP) 1/ has been fully prepared, details of all courses are available, and terms of reference for the foreign consultants have been agreed with the CON. D. General Dam Safety 3.35 Dam safety inspections would be carried out at regular intervals (not less than every five years) for all dams serving NWMP subprojects. They would be carried out by independent Dam Safety Panels (DSPs) acceptable to IDA. The DSPs would (i) inspect the sLructures in detail; (ii) reassess the design flood if needed; (iii) interpret the observations collected since the previous inspection; (iv) check the functioning of all hydromechanical equip- ment; and (v) when necessary recommend modifications or repairs to the struc- tures to improve the safety of the dam. No scheme served by dams with unresolved safety problems will be included in the NWMP. The existing Dam Safety Cells in each of the participating States would compile the data and reports that the DSP would be required to review and would provide administrative and technical support. The costs of the DSP's field visits and other activities would be met from the project funds relegated to dam visits in each of the respective States (para 3.17). Agreements have been reached with GOAP, GOK, and GOTN that they would, by December 31, 1987, establish a Dam Safety Panel with membership and terms of reference satisfac- tory to the Association, for purposes, inter alia, of inspecting peri- odically, and in accordance with sound engineering practices, the dams and related structures serving NWMP schemes in order to determine whether there are any deficiencies in the condition of such structures or in the quality and adequacy of maintenance or methods of operation of the same, which may endanger their safety, and shall promptly carry out any repair works which shall be required to ensure the safety of such dams and structures. Environmental and Ecological Impact 3.36 The marginal environmental or ecological impact of the NWMP will be positive and may include better control of waterborne diseases because of the rotational water supply to be adopted in many schemes and the reduction in waterlogging brought about by minor drainage improvements. IV. COST ESTIMATES, FINANCING, AND DISBURSEMENT Project Cost Estimates 4.01 The project costs including physical and price contingencies, are estimated to be US$157.0 M. This amount includes about US$9.1 M foreign exchange, or about 6% of the total project cost. Taxes and duties included in the total costs are negligible. Cost estimates are based on price levels at the end of 1986. Detailed cost estimates are given in Tables 1 to 6 and are summarized below: 1/ See SDV II, Annex 15, for full details.  -24- Project Cost Summary X of x Total (Rs M) (US$N) Foreign Base Local Foreign Total Local Foreign Total Exchange Costs I. INVESTMENT COSTS A. Scheme Investments 910.7 47.9 958.6 70.0 3.7 73.7 5 64 B. Engineering & Admin. 86.3 - 86.3 6.6 - 6.6 - 6 C. Inst. Strengthening 107.2 4.1 111.3 8.2 .3 8.5 4 7 D. Training 30.7 2.0 32.7 2.4 .1 2.5 6 2 E. Special Studies and Programs 34.3 .2 34.5 2.7 - 2.7 1 2 F. Maharashtra Induction Training 19.8 13.8 33.6 1.5 1.1 2.6 41 2 G. Institutional Support and Training Fund 104.0 26.0 130.0 8.0 2.0 10.0 20 9 Total INVESTMENT COSTS 1,293.1 94.0 1,387.0 99.4 7.2 106.6 7 93 II. RECURRENT COSTS A. Incremental O&M 109.5 - 109.5 8.5 - 8.5 - 7 Total RECURRENT COSTS 109.5 - 109.5 8.5 - 8.5 - 7 Total BASELINE COSTS 1,402.5 94.0 1,496.5 107.9 7.2 115.1 6 100 Physical Contingencies 118.6 6.5 125.1 9.1 .5 9.6 5 8 Price Contingencies 402.0 17.5 419.5 30.9 1.4 32.3 4 28 Total PROJECT COSTS 1,923.1 118.0 2,041.1 147.9 9.1 157.0 6 136 . .... ===== ......=s.su u == === ws= == Physical contingencies average 8%. Price contingencies (282 of base costs) are based on projected annual inflation rates for foreign and local costs of 6%, 6.8%, 6.8%, 7%, 7.1%, 4%, and 4% for the period 1986/87 to 1992/93. Engineering and administration costs are estimated at 9% of the scheme investment civil wovks, which is sufficient to support these minor works by an established organization. Financing 4.02 The proposed credit of US$114.0 N would finance 73% of total project costs (equivalent to about 70% of the water management components and 100% of the special training and technical assistant activities). The credit would cover the foreign exchange component (US$9.1 M) and about 712 of local costs. Expenditures in the participating States would be met from the development budgets of GOAP, GOK, GOTN, and CON. States that might participate subsequently would be required to make similar arrangements. Retroactive financing not exceeding US$5 M would be provided for qualified expenditures incurred after June 1, 1986. The credit would be made available to GOI on standard IDA terms. An agreement would be reached with GOI that itvwill make the proceeds of the credit available to the States on standard terms and arrangements for develop- ment assistance to the States. Agreements have been reached with GOI, COAP, GOK, GOTN, and GOM that by November 30 of each year, commencing November 30, 1987, they would prepare and furnish to IDA for its review and comments, an implementation schedule and proposed budgetary allocations for the following  -25- financial year and, taking into ccnsideration IDA's coments, if any, provide the necessary resources and funds to implement the project. Procurement 4.03 Works under the project would cost approximately US$103.0 M, which includes physical and price contingencies.l/ Although they would be packaged wherever practical, none of the construction activities could be grouped into contracts greater than US$100,000; therefore, none would be of interest to foreign contractors. The contracts will be small because the work involves only minor modifications to many existing systems. The works would be spread over four States and would involve approximately 25 different procurement authorities; the awards would be made over the life of the credit. Contracts estimated to cost more than US$40,000 would use the LCB model documents prepared by the Central Water Commission and the smaller contracts would use a work-order type of document. About 60Z of the works, or US$62.4 M, would be for contracts between US$40,000 and US$100,000. The work-order contracts would be limited to an aggregate of about 20X of the works, or US$20 M. Both types of bidding procedures have been reviewed and found acceptable. The remaining US$20 H of works which are small and scattered would best be accomplish6d by departmental forces. 4.04 Goods consisting of vehicles and equipment would cost about US$2.7 M, which includes physical and price contingencies. Packaging will be done when- ever practical. However, like the works contracts, the goods contracts would be small because procurement responsibility would be divided among some 25 dif- ferent procurement authorities and requirements would be spread over the life of the project. Each contract would be less than US$50,000 and involve only a few of each item; therefore, foreign suppliers would not be interested. All goods would be procured using LCB procedures that have been reviewed and found accept- able. 4.05 Services would be procured on terms and conditions satisfactory to IDA; 2/ the contracts would include about US$3.6 M for assignments to local colleges, institutions, individuals, and consulting organizations and one large consult- ancy package (about US$1.1 M) for the Maharashtra Induction Training Program, which would be of interest to foreign consultants. 4.06 The procurement arrangements and disbursement amounts are summarized in the following table: 1/ Including works on the subprojects (US$101.0 M) and US$2.0 M for civil works under the Maharashtra Induction Training Component. 2/ Consultants will be selected in accordance with Bank guidelines.  -26- Procurement Methods 1/ (USs M) Work Depart- Total Procurement Item ICB LCB 2/ Order mental Other Cost 1. Works 0 61.0 20.0 20.0 0 101.0 0 (42.0) ( (46) 14(.0) 0 T70.0) 2. Goods 0 2.8 0 0 0 2.8 o (1.9) o 6 0 (1.9) 3. Eng. & Adm. 0 0 0 9.3 0 9.3 O O O O U O 4. Incremental O&M 0 0 0 11.9 0 11.9 o b o (8.3) 6 5. Institutional 0 0 0 11.9 0 11.9 Strengthening 0 0 o (11.9) o (11.9) 6. Training 0 1.2 0 2.1 0 3.3 O 1.2 0 (1.9) 6 (3.1) 7. Services, Special 0 0 0 0 3.7 3/ 3.7 Studies & Programs 0 0 0 0 (3.6) (3.6) 8. Maharashtrd 0.0 1.8 0 0.2 1.1 4/ 3.1 Induction (0.0) 1.8 (0) (0.2) (1.1) - (3.1) Training 9. Institutional 0 0 0 0 10.0 5/ 10.0 Strengthening 0 0 0 0 (10.0o (10.0) & Training Fund TOTAL (including 0.0 66.8 20.0 55.4 14.8 157.0 Contingencies) (0.0) (46.9) (14.0) (36.3) (T4.7) (111.9) Percentage of Total 0% 42.5 12.7 35.3 9.4 100 Unallocated ( 2.1) TOTAL CREDIT (114.0) 1/ Figures in parentheses are the amounts financed by the credit. 2/ Prior Bank review would be required on works packages estimated to cost more than US$80,000. These are estimated to cost US$40 M in total or about 20% of total project costs. Goods packages estimated to cost more than US$40,000 would also require prior Bank review. 3/ Many small service and consultancy assignments would be contracted to various specialists. They will be required in the various States and spread over the full project cycle. 4/ Consultants for the Maharashtra Induction Training Program. 5/ Funding for various special project technical assistance efforts would be decided by GOI in the course of project execution. Procurement is likely to include a combination of LCB (for minor enuipment purchase), work order, and departmental or service contracts.  -27- Disbursements 4.07 Disbursement under the credit would be made for: (i) Scheme Investments (a) Works - 70% of approved costs; (b) Goods - 100% of imported equipment and ex-factory cost of local equipment; - 70% for other locally procured equipment; (ii) Incremental O&M Expenditures on a declining scale. For each scheme, during the first, second and third year, disbursement would be at 100X of incremental O&M expenditures, thereafter at zero. This would result, for each state (aggregating the phased entry and completion of the various schemes in that state), in a gradual overall decline in Credit disburse- ments against O&M, at the same time simplifying disbursement accounting. The approved scheme appraisal reports (para 5.09) would define: (i) the estimated baseline of O&M expenditures; and (ii) year-by-year incremental O&M expenditures. Incremental O&M expenditures are defined as the incremental amount over the average provision in the three years prior to the start of O&M support. (iii) Institutional Strenxthening, Training and Special Studies and Technical Services, at 100% 1/ of approved expenditures. (iv) Maharashtra Induction Training Program. Disbursements would be made against 100% 1/ of the expenditures made in accordance with the approved program (SDV II, Annex 15). (v) Institutional Strengthening and Training Fund. Disbursements would be made for 100% i/ of expenditures. Disbursement for works constructed by departmental forces, institutional strengthening, and for payments of less than Rs 300,000 for works and Rs 150,000 for goods itemized by sub-project, would be made against statements of expendi- tures. Documentation of these expenditures would be retained by the responsible irrigation department and made available for inspection by IDA during project review missions. The estimated schsidule of expenditures is given in Table 1. The semi-annual disbursement schedule and the proposed allocation of the proceeds of the credit are presented in Tables 14 and 15 respectively. It is expected that disbursements would be completed by March 31, 1994. The disburse- ment profile assumed follows the standard Indian irrigation schedule. 4.08 To speed up disbursements, GOI has agreed to establish a special account (SA) from which suppliers and contractors would be paid. This SA would be for the equivalent of US$5.0 M. All expenditures from this SA would be promptly reported and documented to the IDA to facilitate its rapid and smooth replenish- ment. The SA may only be used to pay for expenditures eligible for reimburse- ment by IDA. I/ For goods the maximum disbursements would be as in (i) above.  -28- Accounts and Audits 4.09 The project would be subject to normal Government accounting and audit procedures. Although annual audits are carried out in a timely manner, the finalization is a lengthy process involving the presentation of the Statutory Audit Report to the Legislature. Therefore, it would be essential to have an independent auditor prepare an interim certification of the annual financial statements. Such statements certified by the responsible State Accountant General or his designee, supported by an itemized account, a summary of the expenditures, plus the auditor's opinion and comments would be submitted to the Association as soon as possible, but within nine months after the end of each fiscal year. Thus, COI, GOAP, GOK, GOTN, and COM have agreed to: (a) maintain or cause to be maintained records and accounts adequate to reflect, in accordance with sound accounting practices the operations, resources and expenditures in respect of the Project; (b) (i) have the records and accounts for each fiscal year audited, in accordance with appropriate auditing principles consistently applied, by independent auditors acceptable to the Association; (ii) furnish to the Association, as soon as available, but in any case not later than six months after the end of such year, a certified copy of the report of such audit by said auditors, of such scope and in such detail as the Association shall have reasonably requested; and (iii) furnish to the Association such other information concerning said records, accounts and financial statements and the audit thereof as the Association shall from time to time reasonably request. (c) For all expenditures with respect to which withdrawals from the Credit Account were made on the basis of statements of expenditures; Si) maintain records and accounts reflecting such expenditures; (ii) retain, until at least one year after the completion of the audit for the fiscal year in which the last withdrawal from the Credit Account was made, all records (contracts, orders, invoices, bills, receipts and other documents) evidencing such expenditures; (iii) enable the Association's representatives to examine such records; and (iv) ensure that such records and accounts are included in the annual audit and that the report of such audit contains a separate opinion by said auditors as to whether the statements of expenditure sub- mitted during such fiscal year, together with the procedures and internal controls involved in their preparation, can be relied upon to support the related withdrawals.  -29- V. ORGANIZATION AND MANAGEMENT General 5.01 The Operation and Maintenance (O&M) of irrigation facilities is a specialized activity requiring specific skills, experience, and training. In some Indian states, O&M responsibilities are not distinguished from the other responsibilities of the Irrigation Department, basin and project planning, design, construction and quality control, and officers are expected to move from one discipline to another without particular preparation. For example, when an Irrigation Department is part of the Public Works Department, in extreme cases engineers can be transferred directly from road design to canal operation. 5.02 Ideally, the State's would prepare an operation plan for each basin. Allocations among schemes within a basin would be defined and communicated to scheme-level authorities. At the scheme level, an officially constituted Board or Committee, representing concerned government agencies (Irrigation, Agriculture, CAD) and users (para 5.20) would agree to the annual and seasonal operational plan. A single organization would be responsible for carrying out the operational plan within each scheme from the water source to the Government outlet, and for supervising as necessary, the distribution among the farmers below the outlet. 5.03 The organization of maintenance should directly parallel that of operation; thus, the State would allocate funds and program work among projects, and scheme-level staff would be responsible for all works within the project area. 5.04 Under the NWMP, specialization of the O&M organization would be initiated at the state level, and would be fully implemented at the scheme level. The detailed arrangements for implementing the NWMP, and the support to be provided, would reflect individual state circumstances (SDV I, Annex 8), but in each participating state the following would apply; (a) the PPC would prepare NWMP subproject proposals and supervise sub- project implementation; (b) an O&M cell would be established to give direction to the State's O&M activities, and to support and monitor scheme programs; 1/ (c) a scheme-level Committee representing both officials (Irrigation, CAD and Agriculture Departments) and farmers would meet before each irrigation season to approve the operational plan and define the 1/ Initially, this unit would remain within the existing Project Preparation Cell in Karnataka, but in the longer term would become independent. In Andhra Pradesh and Tamil Nadu, it would be established separately ab initio. The balance of responsibilities between the PPCs and O&M cells would evolve over time as the staffing of the O&M cells expands (SDV I, Annex 8).  -30- rules for its implementation, and would meet after each season to evaluate the experience gained; (d) a single O&M agency (either the ID or CADA) would be designated before the scheme investments are completed and would be assigned responsibility for water delivery within each scheme in accordance with the decisions of the Committee; (e) agricultural supporting services would be programmed in the co- .ext of other IDA-supported projects to complement the irrigation manage- ment programs adopted (para 5.18); and (f) systematic linkages would be developed between the O&M agency, other Government agencies, and the farmers (para 5.20). The GOI Water Management Cell (WMC) 5.05 The WMC within the GOI Ministry of Water Resources (MWR) would be responsible for coordinating project implementation and would provide administrative and technical support for the NWMP appraisal committee. This committee would be chaired by the Additional Secretary, Ministry of Water Resources, and its members would be drawn from relevant MWR departments, CWC, the Ministry of Agriculture, and the Indian Council of Agricultural Research (ICAR). An understanding has been reached with GOI that the Water Management Cell would be strengthened and maintained according to a program and timetable satisfactory to IDA. State-Level Organization 5.06 To provide for the necessary State level reorganization and strengthening (para 3.11-3.12), agreements have been reached with COAP, GOK and GOTN that they would establish and thereafter maintain or strengthen PPCs and O&M cells in accordance with a program and timetable satisfactory to IDA. Scheme-Level Organization 5.07 Agreements have been reached with GOAP, COK and GOTM that they would, one year prior to completion of each NWMP scheme, establish and/or maintain a scheme-level committee for each NWMP subproject with responsibility for formulating annual and seasonal operational plans and with representation of agencies and farmers satisfactory to IDA. The Comnittee would comprise officials from Irrigation, Agriculture, CAD and PWD Departments as appropriate, together with farmer representatives. The responsibilities and powers of the Committee would be defined in each scheme proposal. Selection and Approval of NWMP Subprojects 5.08 Primary responsibility for selection, preparation and presentation of schemes would rest with the State Governments concerned. However, GOI clearance would be required to ensure consistency and adherence to NWMP criteria. In this respect, agreements have been reached with GOAP, GOK and GOTN that they would select, plan, design and seek approval of schemes under the NWHP in accordance with procedures, standards and criteria satisfactory to IDA.  -31- 5.09 Schemes selected must have the prospect of improved operation with limited investments of the order anticipated under the NWMP. Schemes would be prepared by the respective State PPC and would be cleared according to the procedures set out in Schedule A. No scheme having any unresolved dam safety issues would be selected (para 3.35). The approved report would form the basis for disbursements under the IDA credit (para 4.07) and inter alia would specify: (i) the area on which disbursements would be based; (ii) the estimated costs of the scheme investments; (iii) the incremental O&M costs, year by year; and (iv) scheme level training plans. Inclusion of sub-projects in States other than Andhra Pradesh, Karnataka and Tamil Nadu would be subject to Bank agreement and clearance by the NWMP Appraisal Com- mittee in the MWR. Munitoring and Evaluation 5.10 Physical and financial progress of the NIMP at the State level would be monitored hy the PPC based on information provided by scK,ane staff. 5.11 At the scheme level, the operational plan would form the basis for monitoring system operations. Regular reports would be submitted to the project committee recording the outcome of seasonal/annual operations (water deliveries, irrigation deficiencies, cropped areas, farmers' satisfaction) and evaluating performance. Taking into account the successes and failures of previous seasons, the operational plan would be modified and operations for the coming season would be defined. Thus, scheme level M&E would be an essential element of scheme management. An M&E plan consistent with manage- ment requirements and manpower availability would be prepared for each scheme and form a part of its O&M plan. The PPCs and/or O&M cells at the State level would help introduce M&E practices at the scheme level, develop report- ing formats and procedures, and monitor progress. 5.12 The PPCs and/or O&M cells would also commission evaluation and other studies from outside agencies to provide an independent assessment of scheme operations or to investigate specific aspects in depth. Taking these into account, as well as the regular reports of scheme staff, the O&M cell would help evolve O&M policies statewide and ensure that information regarding successful and unsuccessful interventions is shared among schemes. The WMC in the MWR would similarly ensure that experience is shared among States. 5.13 The basic guidelines for M&E have been developed (SDV I, Annex 13) and will be further refined. Agreements have been reached with GOAP, GOK and GOTN that they would, by December 31, 1987, prepare a detailed M&E program satisfactory to IDA and implement the program accordingly. Agreements with GOI, GOAP, GOK and GOTN have been reached that they would: (i) prepare and furnish to IDA annual progress reports through the WMC, with the WMC respon- sible for collating State reports and summarizing overall project progress; and (ii) furnish such reports to IDA within three months after the end of each reporting period with the first report to be submitted by June 30, l9P8. In addition, agreement has been reached with GOI that a project completion report would be submitted not later than six months after the project's Closing Date.  -32- Training 5.14 The training program set out above (para 3.15) and discussed in more detail in SDV I, Annex 14, would be organized as follows. Orientation and officer-training seminars and courses would be organized in each State by the respective WALMI, or equivalent, working closely with the PPCs/O&M Cells. Each WALMI would prepare a detailed traizing program and would place an officer in charge of each scheme. He would participate in the initial train- ing programs and would subsequently be available to the scheme authorities for refresher seminars and courses. The WALMIs would also be responsible for developing the training courses and materials to be used in introducing the program to field workers and farmers at the scheme level. The designated WALMI officer would initiate this program, and would provide training and assistance to scheme-level officers, who would subsequently be responsible for farmer training. In large schemes, it is envisaged that a full-time training officer with appropriate support staff will be appointed. In smaller schemes, the scheme officers themselves would be fully responsible and they might be provided with an incentive honorarium related to the scale of the program adopted. 5.15 In view of the considerable administrative and organizational work expected, the PPC would put an officer in charge of training aspects of the project. He woule be responsible for liaising closely with the WALMI con- cerning officer training and for monitoring scheme-level work and farmer training programs as they are implemented in each scheme. Scheme-level training programs would be defined in the scheme reports. For training components that are not part of specific scheme proposals, agreements have been reached with GOAP, GOK and GOTN that they would: (i) by December 31, 1987, prepare and furnish to the Association for its review and comments, detailed schedules for the training program under the Project, and there- after, taking into account the Association's comments, if any, implement such training programs. 5.16 WMC would be responsible for scheduling a program of courses, semi- nars, and study programs for its own staff and for common inter-State programs (including overseas training) for staff of the State O&M Units and selected senior officers in NWMP schemes. An understanding has been reached with GOI that it would by December 31, 1987, prepare detailed schedules for the training program for which the WMC is responsible and, after consultation with IDA, implement the program accordingly.  -33- Maharashtra Induction Training Program (MITP) 1/ 5.17 The MITP will be managed at three levels: (i) a Committee of Direc- tion; (ii) a Management Committee; and (iii) the Board of Studies. The Committee of Direction will include representatives from OI and CO and will provide overall guidance for the induction training policy. The Management Committee will be responsible for the implementation, including administra- tion, financial management, technical coordination, and course content. The Board of Studies will advise on course content and be responsible for project monitoring and evaluation. Agreements have been reached with COM that it would: (i) furnish detailed schedules for the training programs under the MITP to IDA for its review and comments, and taking in consideration IDA's comments, implement th programs accordingly; (ii) by December 31, 1987, establish and thereafter maintain a Management Committee for the MITP with representation satisfactory to IDA; and (iii) prepare and furnish to IDA annual progress reports through the WMC within three months of the end of each reporting period, with the first report submitted by June 30, 1988. Agricultural Supporting Servi:es 5.18 Agricultural Extension (SDV I, Annex 10). A professional agricul- tural extension service, organized on Training and Visit (T&V) principles, has been established with Bank support in each of the States concerned, 2/ and no funds would be provided for financing these services directly under the NWMP. Specific measures would be taken within the context of these extension projects to: (i) ensure that the field-level organization in NWMP subproject areas is fully staffed; (ii) reorganize VEW farmer groups, as far as possible, on an outlet basis; and (iii) appoint water management Subject Matter Specialists (SMSs) in the Districts/Zones where NWMP subprojects are implemented who would contribute to the design of operational interventions and train extension staff concerning the implications for farmer operations. In this respect, agreements have been reached with GOAP, GOK and GOTN that they would: (i) provide appropriate staffing, facilities, services and the other resources required to carry out effective agricultural extension satis- factory to IDA throughout the proposed project districts; and (ii) ensure that such extension services would be fully operational and staffed by the time that project works are completed. 1/ For full details of the management structure and functions, see SDV II, Annex 15 (paras 37-43). 2/ Under the Andhra Pradesh Agricultural Extension Project (Cr. 1219-IN), the National Agricultural Extension Project II (Karnataka Component) (Cr. 1569-IN), and the Tamil Nadu Agricultural Extension Project (Cr. 1137-IN). In Andhra Pradesh, all extension staff under CADA control have been transferred to the Agricultural Department. In Karnataka, 2,000 field staff have reverted to the Department of Rural Development following court action, and although staff shortages will persist, steps are being taken to replace extension staff. In Tamil Nadu, there are no major obstacles, and substantial progress has been made in establishing an effective service.  -34- 5.19 Agricultural Research (SDV I, Annex 9). Clear extension messages for on-farm water management practices are critically dependent on field-oriented agricultural research programs, focussed on the particular problems of each command and responding to the opportunities created by the operational plans. Furthermore, the plan itself should reflect the results of operational research undertaken jointly by the State Agricultural Universities, the DOE, the ID and the farmers. Special subprojects in water management research would be prepared for submission to ICAR under the ongoing National Agricul- tural Research Project (Cr. 1631-IN), which take into account the specific requirements of the NWMP sub-project areas. While responding to the broader requirements of the State, these proposals would be implemented in parallel with, and in support of, the measures introduced under the NWMP and, together with the extension measures outlined above, should help farmers take advantage of the opportunities created. 5.20 Inter-agency Linkages (SDV I, Annex 11). Systematic linkages would be established between the agencies concerned, notably the ID and the exten- sion and research services, to ensure that their operations are coordinated and coherent. Such coordination would be required both at the planning stage ard during implementation. To ensure consistent technical recommendations to the scheme's decision-making body, the staff in charge of scheme operations would meet, prior to preseasonal meetings of the project committee, with their counterparts in the DOA and other agencies. T3 ensure regular interac- tion during implementation, ID staff would participate in the monthly research and extension workshops (at SE/EE level) and fortnightly training sessions (at AEE/Work Inspector levels) established under the T&V extension system. 5.21 Farmer Organizations (SDV I, Annex 12). Informal farmer organiza- tions have been established in many schemes, and existing legislation often envisages more formal arrangements (e.g., the 1984 Andhra Pradesh Irrigation Utilization and Command Area Development Act). Once a reliable water supply is provided, such organizations can be expected to operate more effectively (para 2.18). They would be promoted in NWMP schemes on an outlet/minor/distributary basis by the ID/PWD (in Karnataka, in association with CADA; see SDV I, Annex 8). Initially, such organizations would con- centrate on water distribution and maintenance below the Government outlet and would be a mechanism for systematic interaction between the farmers and the project authorities. Ultimately, they might be fully integrated into the management of the scheme (e.g., through elected representatives on the scheme committee). The speed with which this can be achieved would depend on the effectiveness of the operational plan adopted and would vary from scheme to scheme. VI. AGRICULTURAL PRODUCTION, FARM INCOMES AND COST RECOVERY A. Agricultural Production Introduction 6.01 The project would increase agricultural output primarily through more effective use of canal supplies. Changes in the timing of irrigation issues, in the methods of water delivery, in the efficiency of water use, and in the effectiveness of supporting services would by themselves and together modify  -35- cropping patterns, increase irrigation intensities, and raise yields and value added. These changes should also ensure more effective utilization of groundwater resources. They would not normally augment canal supplies, although this may occur when reservoir operations are modified or reservoir, tank, or canal carrying capacities are increased. 6.02 The relative importance and effect of different interventions would vary widely. In many cases, higher production can be achieved as a result of operational changes alone. In all cases, however, higher production would be consolidated and augmented by the physical works to be financed under the project. To illustrate the likely production impact, three schemes have been evaluated in detail: the Rajolibanda Diversion Scheme (RDS) in Andhra Pradesh, the Vanivilas Sagar Scheme (VVS) in Karnataka, and the Sathanur Scheme (SS) in Tamil Nadu. No allowance has been made for any impact on groundwater abstractions, although this could be significant, particularly in the case of he Sathanur Scheme. Nor have the possibilities for augmenting canal supplies been taken irto account, although this can be envisaged in all three ca-ts: in RDS if the capacity of the main canal is increased under a second phase, in VVS as a result of improved utilization of drainage run-off at the downstream Narayanapura Anicut, and in SS if the rules governing releases for the downstream ayacut (the "old" ayacut) were modified (see SDV I, Annexes 5, 6, and 7). 6.03 The projected increases in production represent an assessment of the project's direct impact and do not represent the full long-term potential. With experience, further changes may prove desirable, possibly with the support of additional modernization investments. Higher field application efficiencies, increased yields, and/or the the spread of high-value cropping may also result from parallel agricultural research and extension programs or from farmer response to market forces. Thus, although a systematic approach to operations is a prerequisite for achieving a scheme's full potential, it represents only one step in a continuing process. Impact on Agricultural Production in Selected Subprojects 6.04 Rajolibanda Diversion Scheme (SDV I, Annex 5). The principal fea- tures of the proposed operational plan are to: (i) preclude paddy in middle and lower reaches in kharif and throughout the command in rabi by introducing systematic intermittent irrigation; 1/ (ii) stagger kharif cropping by sub- command down the main canal; (iii) constrair the period of cropping and encourage short-season varieties by adhering to pre-announced canal closure dates; and (iv) support cotton cultivation in blocks where it has been estab- lished. These objectives would be supported by investments designed inter alia to improve system control, modify canal capacities as required, reduce canal losses, and rehabilitate the channel system. Given the history of the scheme and its severe water shortage, it is unrealistic to expect water to be spread equitably throughout the command. Rather, the aim vould be to serve the existing irrigated areas more effectively. Thus, in rabi, the area 1/ Even if this objective is generally achieved, limited areas of paddy would probably continue in areas affected by salinity and/or waterlog- ging.  -36- served would be reduced in order to avoid repeating past failures. In kharif, the area can be extended by limiting paddy and, if AP can reach an agreement with Karnataka to remodel the main delivery system (Stage II of the proposed plan), it might be possible to extend the kharif area over most of the CCA. Table 18 summarizes the expected impact on yields, cropped areas, and production. 6.05 Vanivilas Sagar Scheme (SDV I, Annex 6).3 The principal features of the proposed plan are to: (i) carry over 375 Mm of stored water to guaran- tee the perennial garden crops and ensure a full irrigated kharif crop over an increased authorized command (from 11,000 to 14,000 ha); (ii) provide water for rabi only if a sufficient amount is available for the full increased command (estimated to occur 2 years out of 10); (iii) adapt the intermittent schedule to respond better to the requirements of ID crops (by reducing the irrigation interval and responding to peak demands); (iv) con- centrate paddy in low-lying areas and discourage sugarcane cultivation by not reopening the factory; 1/ (v) maintain steady runoff flows at the downstream Narayanapura Anicut by alternating operations of the LB and RB canals; and (vi) substitute direct canal supplies for those now served by pumping return flows. These and other measures would be supported by rehabilitating the canals and structures, providing new control structures, extending the dis- tribution system where necessary, providing field channels/drains in the extended command, and lining vulnerable sections of the main and distributary canals. As a result, some improvement can be expected in water use efficien- cies: the area under high-value garden crops would continue to increase; and kharif cropping would be assured. Since the garden crop area would expand even without the project, Table 19 summarizes the expected impact on cropping with respect to average conditions of the past ten years and to projected future "without" project conditions. 6.06 Sathanur Scheme (SDV I, Annex 7). The main features of the proposed plan are to: (i operate the left and right bank commands on a common basis, opening the canals in mid-December to serve the full area if storage is available and a limited area in poor years; (ii) provide water on an inter- mittent schedule, and with a closing data responsive to ID requirements, thus preclude paddy on the direct ayacut except in low-lying areas with supplemen- tal groundwater; (iii) provide direct supplementary supplies to the system tanks to allow a full paddy crop during the NE monsoon, utilizing tank storage to exploit surplus flows in good years; (iv) provide supplementary irrigation during kharif to support ID cropping during dry spells to the extent allowed by carryover and summer flows; and (v) review the rules governing releases for the downstream ayacut and for the operation of upstream storages, within a river basin context, with a view to developing revised rules for optimizing water use. These and other measures will be supported by investments designed inter alia to improve system control; rehabilitate canals, tanks, and structures; modify canal capacities; extend the distribution system as required; ard provide for direct supply to the tanks. As a result, some improvement can be expected in water use efficien- cies: in about 50% of all years, there would be a full post-monsoon ID crop 1/ The factory would require considerable capital investment to rehabilitate.  -37- together with supplementary supplies during the monsoon equivalent to three irrigations; in about 252 of all years, there would only be a post-monsoon crop; and in the remaining years there would only bt a restricted post-monsoon crop. In most years, there would also be a full paddy crop under the tanks. Table 20 sumnarizes the expected impact on yields, cropping patterns, and production. Conclusions 6.07 These three schemes indicate the likely impact of the NWMP on agricultural production. In general, the project is expected to have the following effect: (a) a decline in paddy, and perhaps sugarcane production as high water-use crops are replaced by those utilizing water more effi ciently; (b) a marked increase in the output of oilseeds and pulses; (c) a limited impact, either positive or negative, on the output of coarse grains depending on the relative importance of the substitu- tion of rainfed by irrigated cropping; and (d) a positive, and perhaps substantial impact on high-value cash crops such as cotton and coconut. Crop diversification may be greater than suggested above, in particular if there is a significant impact on groundwater extractions, with vegetables, farm forestry and horticultural crops all being promoted. Furthermore, while crops grown on residual moisture may be affected adversely by declining paddy cultivation, new opportunities should be created following irrigated kharif crops and through better control over the timing of crop cultivation. While paddy may decline overall, it would still predominate below system tanks and in low-lying areas where seepage precludes other crops, and would continue to have an important role in the overall cropping system. 6.08 Assuming these three schemes are representative of the NWMP as a whole, then the total impact of the project might be in the order of the following: 3 Schemes NWMP as a Whole (67,000 ha) (625,000 ha) (tons) Paddy - 7,770 - 72,500 Groundnuts +19,980 +186,400 Sorghum + 2,870 + 26,800 Millet + 6,220 + 58,000 Sugarcane +20,510 +191,300 Cotton + 1,980 + 18,500 Coconut (copra equiv) + 2,660 + 24,700  -38- B. FARM INCOMES Market Prospects and Prices 6.09 Incremental production would be a small percentage of South India's total agricultural output and could be readily handled through existing marketing channels. Prices of foodgrains have adjusted to the marginal surpluses that India has achieved in recent years and are expected to remain more or less at their present levels in real terms. Prices of groundnuts and other oilseeds are also likely to remain fairly steady, despite an expected decline in world market prices, given India's high income elasticity of demand for edible oils and GOI's continuing efforts to reduce its large vegetable oil imports. Cotton prices will be determined primarily by the world market, although Government intervention through tax and other measures tQ assist the textile industry is likely to keep domestic prices below those on the world market. In contrast, sugarcane prices have tended to exceed the import equivalent price, except during periods of world shortage. Given the project's widely dispersed impact and the modest increases in production to be expected in any one location, no significant impact on domestic price levels or marketing channels is envisaged. Farm Incomes 6.10 With the introduction of improved irrigation plans, the distribution of water to farmers can be expected to change--usually the volume of water used at the head decreases, and that at the tail increases. Experience thus far is that this does not cause great difficulties: first, because the planned pattern of distribution is published, the equity of the proposal is clear to all concerned parties; second, a major cause of overirrigation is the desire to insure against uncertain supplies, but the plan of operation will greatly reduce uncertainty. Farm incomes vary enormously, depending on the size of farm, the crops grown, and the irrigation service provided. To illustrate the direct impact of the project, representative farm budgets have been compiled for the three schemes appraised. The results are summarized below and the details given in Table 28. Family Income Size of (Incl. return to Family Labor) Farm Without With Increment (ha) _IR9) (Rs) (Rs) Rajolibanda Diversion Scheme Rainfed Farm 2.5 2,405 2,405 Kharif Farm 2.5 4,320 7,620 3,300 Rabi Farm 2.5 7,380 11,980 4,600 Cotton Farm 4.0 21,075 31,700 10,625 Average (Weighted) 2.5 4,955 8,215 3,260 Vanivilas Sagar Scheme Average Farm 3.0 17,190 22,410 5,220 Sathanur Scheme Average Farm 1.5 5,535 7,255 1,720 6.11 In the RDS, irrigation intensity is expected to remain below 70% and the major objective is to serve this limited area in a more reliable manner (para 6.04). As a result, incomes are expected to rise by about 30%, with hieher incomes accruing to farmers receiving rabi rather than kharif supplies  -39- since they have the potential to obtain enhanced yields (notably of ground- nuts) as well as an additional rainfed crop. In practice, some head-end farmers obtain both rabi and kharif supplies whereas many tail-end farmers remain dependent on subsistence rainfed farming. Cotton farmers tend tc have larger farms and to obtain higher incomes in return for heavy cash outlays. Following project implementation, the above estimates--which are based on an average family size of six--are equivalent to about US$405 per capita for irrigated cotton farmers, about US$155 for rabi farmers, US$100 for kharif farmers, and US$30 for those dependent only on rainfed farming. 6.12 The objective in the VVS and SS is to distribute available supplies in an equitable manner throughout the command (paras 6.05-6.06). Although farm sizes and cropping activities vary, most farmers have both garden and field cropland in VVS, and both paddy land under a tank and higher land in SS. In both cases, therefore, budgets have been estimated for an "average" farm based on the overall assumed cropping pattern and relative farm sizes. The higher incomes projected for VVS reflect the prevalence of high-value garden crops and larger farms and, following project implementation, are expected to reach the equivalent of about US$285 per capita. In SS, the corresponding level is only US$95 as a result of the greater farm population, although those farmers with a well can expect irrigated crop intensities, and hence farm incomes, to be substantially higher than the projected average of about 154% (Table 20). 6.13 The estimates presented suggest that average farm incomes may rise by about 25-33% as a direct result of the NWMP. This provides a reasonable ;iLcentive for ensuring adequate farmer response to the opportunities created by the Project. Since it is still assumed that yields will remain well below their theoretical potential, the longer-term impact of the project may be considerably greater. Without a reliable and predictable water supply, there is little prospect of realizing this potential but, in order to do so, other complementary agricultural research, extension, and related measures are also required (see SDV I, Annexes 9, 10, and 11). C. O&M FUNDING AND COST RECOVERY Background 6.14 It has been the policy of GOI and most states that water and water related charges should cover O&M costs and a proportion of capital costs taking into account farmer repayment capacity. However, at issue are: (i) what would be an appropriate level of O&M; and (ii) the most effective recovery methods given the inequity and unreliability of water supply in most irrigation projects, the administrative constraints and the political environment. O&M Funding 6.15 There is a general awareness in India of the need to improve the level of maintenance. However, the way of achieving this objective with least cost and on a sustainable basis under varying conditions is not entirely clear and needs additional studies and investigations. It is generally felt that the share of O&M allocations spent on actual maintenance works is too low and that additional allocation of funds to O&M could be  -40- absorbed by staff costs unless accompanied by reorganization of O&M opera- tions and formulation of standard O&M procedures. Another issue related to O&M is the possible involvement of water users' groups (WUG) in both opera- tion and maintenance which could reduce the part of O&M costs borne by the government. Once main system supplies are provided reliably, farmers can be expected to assume greater responsibility for minor system O&M (para 2.18). However, while success with WUG has been rather localized, their active involvement on a large scale is being tried throughout India and some sig- nificant results can be expected within the next 2-3 years. The method of system operation will also affect the O&M costs; for example, a system with minimal number of gates operated according to the warabundi principles (para 2.06) would cost less than a system operating under Shejpali (para 2.10) as manpower requirement would be smaller. The implementation of the NWMP sub- projects would provide valuable experience with improved O&M planning proce- dures at the scheme level. 6.16 While GOI and the Stntes are genuinely interested in identifying and implementing viable O&M solutions, the discussion above indicates that the problem is more complex than estimating a statewide figure for O&M costs. Increasing financial allocations to O&M is not likely to be sufficient without concurrent solutions to the complex management issues. The bulk of such increase could result in an enlargement of an already excessive estab- lishment, rather than be spent on much needed maintenance works. 6.17 To enable the formulation of a viable O&M policy for each state, the project would support the following: (a) a properly staffed state-level O&M cell for overall direction of O&M activities, as well as planning, implementation and monitoring of scheme level O&M programs (para 3.12); (b) the incremental O&M costs in NWMP schemes would be partly funded --on a declining scale--under the project, with the respective States committed to a scheme-specific level of O&M activities, which would be closely monitored; and (c) an. interim O&M program review would be carried out, the purpose of which would be to review initial progress in the pilot schemes and develop detailed plans for further studies and investigations which the project-supported O&M cells would then extend, as appropriate, to other State irrigation systems. Agreements have been reached with GOAP, GOK and GOTM to: (i) under terms of reference satisfactory to IDA, carry out a detailed interim operation and maintenance review of the Schemes; (ii) by December 31, 1991, in consultation with IDA, review the results of the review; and (iii) taking into account such review and the operation and maintenance experience gained in other irrigation projects elsewhere in each Participating State, not later than December 31, 1992, in consultation with IDA, prepare an appropriate Action Plan for improving operation and maintenance practices in each of the Par- ticipating States, and, promptly thereafter, initiate the implementation of such Action Plan.  -41- Cost Recovery 6.18 General. The main sources of cost recovery from irrigation projects are: (i) direct water charges; and (ii) indirect charges, mainly a sales tax. Changes in the level of water charges have to be approved by the Legis- lature in each State. For administrative reasons, water charges in most states are usually uniform throughout the State, by crop and season. Most of the States have difficult recovering costs through water charges since water supply is unreliable and inequitable in many irrigation systems, and there- fore, farmers are reluctant to pay water charges. Consequently, a con- siderable part of the recovery is indirect, through the sales tax on marketed output. 6.19 Improvement in irrigation services of the type proposed under NIMP would improve cost recovery in two main ways: (i) improved collection of water charges--experience has shown that where water supply is reliable, farmers are more willing to pay; and (ii) increased state revenues from a sales tax on part of the incrementaL production resulting from such improve- ments reaching the marketplace. 6.20 Cost Recovery in Participating States. Direct charges for irrigation are higher in Andhra Pradesh than in Karnataka or Tamil Nadu, but are still fairly low, and the procedures for their assessment and collection are generally inadequate. In Andhra Pradesh and Karnataka, betterment levies have been successfully challenged in the courts, and even in Tamil Nadu it would be extremely difficult to levy such a charge under the NWMP, given the project's limited physical improvements. The most important sources of cost recovery--notably the sales tax on incremental agricultural output--are likely to be indirect. In addition, the rate of collection of direct water charges should improve. 6.21 In all States, a water-related charge is levied on publicly operated schemes, with the receipts entering the Government's general revenue account; however, farmers are eligible for remission if water is not received. The system and levels at which charges are levied differ from State to State: (a) Andhra Pradesh. Both assessment and collection are undertaken by the Revenue Department, and the charges discriminate among crops, categories of irrigation service, and first and second/third crops. Rates were increased by more than 50% in March 1986 as follows: Category I a/ Category II b/ 1st Cro 2nd/3rd Crop lst Cro 2nd/3rd Crop ------------Rs/ha) c--------------- Paddy 148 222 99 148 ID Crops 99 148 49 74 Two Seasonal 371 247 aT Good and dependable sources of irrigation providing assured supplies for 8 months or more. b/ Less dependable supplies available for less chan 8 months. c/ Water charges are raised on a per acre basis.  -42- (b) Karnataka. Assessment is made by the Irrigation Department--in principle, it is based on a joint inspection with Revenue offi- cials--and collection occurs through revenues. Various Government Orders have been issued revising rates (in 1979, 1980, and 1985), but in the absence of legal sanction, collection continues at the levels established in 1976. The 1985 proposals will become eff;ective once the Irrigation Rules have been amended, probably by 1988. They compare with the 1976 levels as follows: Established in 1976 Proposed in 1985